Couple things: Roll center height alone is not a metric, it's in relation to the CG height. Yes spings and anti-swaybars control roll rate as well but then you get into ride quality. Also, roll centers need to be looked at in pair with the rear, creating a roll axis. This will tell you where weight is transferred. If the front RC is high but the rear is low, you'd get weight transfer on the outside rear, potentially lifting inside front and vice versa. Most auto manufacturers design cars to understeer to get drivers to slow down. This can be done with roll rates (springs, swaybars), instant centers, or camber rates. So take those dimensions with a grain of salt and look more at the apsects of handling these cars have with what your going for and how the weight is balanced. Before you finalize control arm and mount geometry, you will need to figure out where to put the steering rack. You are limited by its mounting height, but more importantly its length, and you may have to compromise on camber gain or RC height to minimize bump steer.
Great points. I'm not sure how to put this all into one coherent video. I'll do a little better with the RCH description as I know I oversimplified that point. And it's an important point. I did design with the steering rack in the loop - so from the front view, it's installed in the model and there is a bump steer output based on the FV only. It get's left alone until I do the top view then it's set. I have the bump steer dialed out at this time and when I put it into 3D then that parameter is (and this will start an argument) is probably the most important.
@@ThrottleStopGarage I did forget about being concise for the video. Bump steer is definitely important and I'd focus on it more than camber gain. Going forward, I'd say also say bump steer is more critical than ackerman. You have slip angle regardless (just how tires work) and that slip angle changes depending on sidewall, tire pressure, intial camber, and tire load in the corner (so springs and swaybars). You dont have to be super precise with a numeber, but there is a range. Iirc the c5 corvette is ~130-140%. As a reference, my awd s10 (lowered and low pro tires) has the factory geometry of around 180% and the 320tw tires fight each other in parking lots.
Roll center axis does change lateral load weight transfer distribution yes, but does not change diagonally. Meaning, in the case of high rear roll stiffness, the outside rear tire may increase it's load but the front and rear weight distribution remains the same. Both the front tires loads could remain the same, but the inside rear decrease as much as the outside rear increases. To have longitudinal weight transfer, there must be longitudinal acceleration. It's a common misinterpretation for people getting into vehicle dynamics Also, roll centers are fake, but that's a deeper dive 😅
@sevenducecut the roll centre debate does my head in...I understand the desire to get a reference location, but the nature of the beast can't really be what is described. If you have specific advice...I'm all ears.
@ThrottleStopGarage You did your research and did just as well as anyone getting into this black magic. Roll centers are a simplification of forces going through the control arms. The calculations of roll stiffness using roll centers are useful when you have a suspension system already made, and need to see what the system does as you make adjustments (on a car with adjustable points). A more complicated but closer to truth method uses instant center position to find jacking force, and IMO is better when first designing a system from scratch. I haven't seen this method in books, so I've had it on a video to do list for months 😅
I have watched you working this project for a while now, and several reiterations on the front suspension. Your explanation is actually making sense. Thanks
@@ThrottleStopGarage my step-father had a '64 AC Cobra with the leaf spring front suspension, I was going through the suspension while you were explaining it.
@@ThrottleStopGarage The front suspension was very similar to the AC Ace of the '50s, transverse spring acting as the upper control arms. Very long UCA's with the inner pivot functionally mobile during its travel. Might be time for some plywood mock-ups.
When I was in service, doing hundreds of maintenance and crash alignments over my 35 years, I kept an RC car with adjustable A-Arms, to demonstrate what you're showing. It had tiny bubble levels glued all over it. Great visual tool.
Cool - I've got a little model I made to prove the Ackerman relationship to steering rack position. I just couldn't visualize the complex relationship.
@@ThrottleStopGarage Are your "adjustment range" calculations accounting for vehicle loaded weight, driver and full gas? I'd be bummed if I ran out of caster or camber(ride height dependant). And trying to nail the spring rates will be added rocket science. Generally speaking, alignment spec loads are with 3/4 fuel and a 70kg driver, which I guess is out of date now. eyeroll.
@@kathysarmcandy1992 That's the idea...I've got several spreadsheets running with spring/roll and other calculations. I'm not sure how ultimately useful these are as they're copied formulae from others (or were donated by friends who have developed them). My guess is that they're close enough for a first approximation. It will need to be tuned when rolling.
@@ThrottleStopGarage Adjustable upper ball joints, I see similar to your screw-ins. Don't forget to make friends with an alignment shop. On and off hours after work and buy the shop pizza is totally worth it.
Xf motorsports did great videos on suspension design a few years back. The guys are super responsive over there too and helped me with my trans adapter plate design. Ps they are also canadian
Really appreciate the detailed breakdown of suspension dynamics. I've run into the same problem while trying to learn this; every resource I found was far too basic. The order of operations when designing a suspension from scratch is a game-changer. Thanks!
14-15 year old me (a third of a century ago, now...) designed a car, and I had that same wonky control arm setup like your C4 Corvette suspension on the bench. My single source for guidance (from the public library) wasn't clear enough to me, and I tilted the top arm so the pivot on the top of the knuckle was lower than the pivot on the frame. Anyone roughing this out -- this was all pre-internet, so I was lucky to have a small town library with just that ONE book at the time. Several years later, re-looking at my drawings in CAD, I mapped out the virtual center, and saw that camber curve: it was dished concave in, not convex out how we normally want for camber GAIN. That was my aha-moment to tilt the top arm the other way. We got stumped both by the camber curve. Thanks for putting me back on track! Your explanation did clear up some of the murky less understood design process past where I stopped in my teens. The visuals that Racing Aspirations provides helped backstop all the concepts better, as you explained. It probably helped that I had some, not all those concepts percolating in my head now for 30+ years after my first and only go at a suspension design, so as you explained it on video, I wasn't falling behind. Very interested to see what the next step holds, how the 3d-analysis is done. Onto the voodoo and dark arts of anti-dive, anti-squat, and more! And, thank you again, for taking us along for the ride. :D
Thanks - it took a lot of work to get these concepts nailed into my head. I literally had to read and take notes and go out to the garage and play with things until how to design instead of analyse. I'm still not confident.
@@ThrottleStopGarage I get the slow grok on the concepts. I think you explained things well, especially all your emphasis on the things that were murky to understand from all the technical reading. The only thing that seems unclear to me, just because of that oddball upper control arm geometry on it as it is right now: Was the C4 setup like the one on the bench, really? with outward/downward sloping upper Control arms so the camber curve was all camber loss? Or was this C4 setup modified by the seller in ways that made it set up so? :o
@chrstphrr the outward sloping UCA is how it was designed. That's how it is in the drawings and on the bench it is set so all points are located as they would be at ride height. There is another set of mount holes in the cross member that are lower. I've never seen anyone mount the dog bone down there, but it's tempting.
Well that certainly explains why I’ve always felt I was spinning round and round in successively smaller circles when I’ve tried to understand suspension design. Reading the books made my eyes glaze over. Sometimes after giving up I would even head for the donut shop for a superior glazed solution. Unfortunately that never helped either. You’ve inspired me to give it a go. Definitely looking forward to the next episode. In the meantime I’ll be rewatching your previous episodes and reading all the comments to help drive this home. I’ll probably even pull those dusty books back off the shelf for another spin. Thanks so much Craig for slogging through this. 🙏🙏🙏🙏
Thanks - this was a game changer for me. Like you, I could read the words and even explain it, but the shapes wouldn't work in my head. I had no "sense" of things like the viewer who pointed out the problem in a suspension by looking at it. Gold star for him. This did come to me like a clap of lightning. Boom - suddenly it all snapped into place.
Don't measure from the tip of a tape measure if you care for good accuracy, "burn an inch" or perhaps 10mm in your case and measure using the divisions and not the fairly approximate tip of the tape :) The effort you're going into for this is pretty impressive and your willingness to jump into areas where you know you need to learn is great! I will admit the suspension stuff is Greek to me although I've certainly heard many of the terms and have some vague understanding. Much appreciation for trying to explain it as you're learning! The feedback you'rew getting is also awesome and I cannto wait to see what you end up building and how you like driving it eventually afterwards. It's like you're building Tally Ho in automative form - I love it!
I know you already replied that the level was not placed as expected on the hub, but I think it’s important to notice that you were reading the “bubble” level backwards. As you increased bump on the C4 suspension mock up, camber was going more negative. I watched a promo video for the tool to confirm. I don’t mean that this is the right setup for your car, but I don’t want you to be confused and frustrated in the future when using that tool. Also, you are embarking on a long journey if you plan to fabricate all the components. Mike Burrows from the channel Stance Works did something similar and re-did it a few times. He has some good videos to reference on the fab techniques and learnings. Good luck!
Thanks - I noticed that after the video was released. I had moved the level to shoot the segment. It happens. I understand the nature of the journey - looking forward to most of it whatever happens.
I'm glad I found this video. I do want to do a swap as my car has a solid axle type and parts are becoming rarer. A custom suspension design would allow me to build it with easier to find parts.
Talking about camber gain around 9min. Actually looks like you went from a half degree postitive camber to negative 1.5 with bump (i assume your angle gauge 's bumb feature should be same as a level bubble). Regardless of that, its still not enough static camber at ride height or enough gain through the travel. Such a low roll center is counter to where I'm going through with testing on my car. However yours is WAY lighter than mine. If you have the room don't be afraid to add chassis side adjustment for the lower arm that's higher up. Which will directly raise the roll center. Up and down adjustment for the upper arm is a good idea too. I was once told anything that can be made adjustable is a good move. The camber gain target you showed seems too little if you plan spirited driving. Super basic but i set mine by sitting on the fender and using the angle cube to measure body roll (2.4), then set the cube on the tire to check for gain or loss. I went with a degree more negative gain than body roll as a place to start. Best year I've had yet in tire longevity/wear doing street, autocross, and track days. I don't know what that gain is vs compression travel but it is definitely more than .6/inch. However a street only setup probably wint need the additional camber gain. In aggressive driving it does need more because you also have to compensate for tire deflection. Overall, good video! I also spent months researching just to start understanding what's all going on when everything moves.
When I filmed the level, it wasn't flat to the hub and I didn't notice. The computer model is correct and it's not great. I am still working on the roll and camber curve. This will get more attention as I do the side view.
@@ThrottleStopGarage Oh, and if you haven't figured it out yet. The lower arm primarily locates the roll center height and the upper dictates camber gain. Yes, both affect on the other but to a much lesser effect.
@@ThrottleStopGaragesomeone else was commenting about the rack placement. Easiest/obvious placement is with the tie rod pivots lined up with the lower control arm pivots. Then place the outer tie rod lined up with the ball joint. If they follow the same arc then it can't bump steer beyond some pretty minor differences.
Thank you Craig for doing what few have the patience for. Basics, camber gain; when you "bumped" the wheel up first 1" and then 3", you lost negative camber gain, you didn't gain camber. Maybe you just misspoke there but it's an important point when we're talking negative vs positive, losing vs gaining, concave vs convex bla bla. The upper control arm is obviously shorter than the lower, as both move through their respective arc ranges, the shorter arm has no choice but to result in offering "negative camber gain". The problem is the upper and lower arms are not parallel, causing some "loss" at different points through the range of travel, but ultimately will give negative camber gain. The geometry has no choice but to produce this result, the upper arm is shorter. In the Racing Aspirations software it is clearly set up differently to what you have on the bench in your garage, in the software version the arms are very nearly parallel, the Corvette is very clearly not parallel. Sorry if this sounds like I'm angry, really I'm not, I'm only wanting to clarify why you lost camber gain, the upper arm begins, lets call it drooped, the lower arm has some degrees of "rise" from the "at rest" starting positions of your experiment to see what the suspension was physically giving you. This is where the dark arts show up, bump steer, weight jacking, roll center, roll axis on and on, this is why this is all so difficult to master. Before you move on to the side view, you have more work to do from the front view. Ford went to an extreme to "correct" for roll center migration, they designed the infamous Twin Torsion bar suspension. This provided for a really smooth soft supple ride, but did nothing for handling or tire wear. It was a cheap solution or alternative to Chevy's double A arm suspension, which is as you are finding out, very complicated to mitigate all the negative compromises required for mass produced suspension. Remember the Corvair, lawsuits were abundant because the geometry/design was flawed from the beginning. As important as reading and studying the available printed material is, like another "commenter" said, an RC, radio controlled toy car can teach you as much as all those books combined. I highly recommend getting a RC car to test your theories/setup/geometry. Don't get me started on roll center, every race car mechanic has their own opinion of what is the optimal height and they have tested their theories on the track, go to a local circle track and talk to those guys, invaluable. Pavement or asphalt racers would be optimal as you want your car to run on the street. Oh, and potholes, don't forget our Alberta roads mostly suck.
Thanks - I did fumble with the words a bit there and the level was hung on the edge of the lug nut and showing the opposite of what is happening at ride height - that segment is a bit of a mess. Viewers caught it - I'm too tired. The computer model is correct and the path is funky for the C4 - good enough and I can sleep well enough.
Cool video! I use the C4 front suspension for my race car, and I did a few things to help (not solve) the issue that you observe. My first solution was to get a longer lower control arm ball joint from Howe. Does put the control arm close to wheel hoop though, but 18" wheels clear fine. I also chucked the upper control arm cross bar spacers in the lathe and cut them down to the absolute minimum in the back and enough in the front that I could run more caster than stock for camber gain under steering input. Still doesn't give more than bout 2 degrees maximum static camber and I am going to go to a custom shorter upper control arm to get that sorted. Lowering the upper arm inboard pickup points would help as well, but not sure how much fabrication I'm willing to undertake. Nice video, thanks for making it!
When I was autocrossing my C4 ages ago, I put offset lower control arm bushings in which increased the static negative camber markedly. These are still readily available.
FDF RaceShop in Ontario makes angle kits for drift cars (aka a lot of experience with non-OEM front suspension geometry modification). I believe they have their own engineering department too. Might be worth trying to reach out to them… they’d probably be atleast a little interested in helping out upon hearing you built a 60’s Volvo body out of carbon fiber Great video. Really appreciate the visual representations! Probably a huge PIA to film and record
This is incredible. You did a great job of explaining! I feel like anyone who is doing the same will save TONS of time and money by watching. Remember, if the women don't find ya handsome, they can at least find ya handy.
I really need this. I am very reluctant to continue ‘trying’ to work with 40+ year old suspension parts and constraints, especially when it is well known that bump steer was an issue back then. I actually like the idea of starting from scratch on my mk7 Bolwell.
Great video and a good attempt or start at solving your problems. I am not aware what tires you plan on running. If they have a larger side wall and will have significant contact patch migration under cornering load, then you may need to increase the scrub radius. You want to insure that the forces generated at the contact patch stay on the same side of your SIA as you load it up dynamically. You don’t want to have the forces at the contact patch go past the SIA, as when this happens any feedback your feeling in the steering wheel will also change direction, pulling the steering wheel in the opposite direction mid corner. Next thing, as the SIA increases, you get more / less camber gain as you turn the steering wheel. So if your benchmark cars were getting 0.6° camber gain, but had a lower SIA angle, you might need to design in more or less camber gain in your wishbone length ratios, to accomodate for the camber loss you get when turning the wheel with a high SIA angle. Looking forwards to the next video. Cheers. Ryan.
Thanks Ryan - 225 45 17's are the target tire - probably a Michelin Pilot Sport 4S or similar. Is there a source that I should read to take a stab at ensuring that the scrub is in the right range for the tire. I've never been able to squeeze tire data from the MFG's (and I've tried). I wonder if there is a simple tire model that would be good enough for a first approximation?
@@ThrottleStopGarage I have not looked. I know Avon motorsport did have some tire data available for their motorsport tires. You might also be able to get data from some of the other motorsport tire suppliers. I've never tried. I think as long as we are aware of it, and have some idea about where it will be, then ultimately, we can design for it. I'm going to assume you have seen some GoPro type footage somewhere that shows the tire being stretched, pulled across from its static position and heavily loaded mid corner? If we use that as our crude basis for estimation we can then say we assume that the tire side wall goes from being vertical/perpendicular to the surface, to say at absolute worse case, 60° from perpendicular. If we know the side wall height (difference between tire radius and rim radius) then we can calculate how the contact patch may migrate laterally under load. All I am concerned about is keeping this contact patch on the side of the SAI during cornering. There is also a side view component to this as well. Caster and hub offset forwards or rearwards of the SIA come in to play in this orientation. My thinking is that making sure the contact patch is in a known position relative to the SIA is still valid to keeping a consistent feel in the steering under all conditions. If you go back to the C4 corvette. The tire size was 255x45 R17. So a side wall height of 255*0.45 = 114.75mm? (is it radius or diameter?) * Sin 60° = ~99mm. IE under load the stock Vett front contact patch may move up to 100mm or 4" assuming the 114.75mm was the radius. Hence why its knuckle offset may have been where it was.
The explanation of front view design is great, start with the track width you need, add a few locations at the wheel, start from the center of the car then connect the dots. I'll be waiting for the side view and am hoping for a top view vid as well. Concerning top view, most if not all street cars ( use a 77 to 90 full sized GM ) have the lower control arm rotated so the front pivots are closer together. I'm not sure why this is done but I think it interacts with what I'm typing below. The upper is rotated as well but I'd have to look at a suspension to see what way. As for side view, the control arms pivots are tilted relative to each other , I'm pretty sure this has to do with anti dive where braking forces try to add spring rate to the front suspension. Think of it this way. If the brake was locked and you applied a torque in the direction of forward motion , how much torque would it take to move the suspension into droop? High torque = minimal anti dive = nose droops a lot under braking. Low torque = high anti dive = minimal nose droop = possibly too high of an effective spring rate leading to turn in problems.
Thanks Throttle Stop Garage ... very, very well thought through and articulated ... about the most helpful suspension geometry video I have seen. As I am building a GT40 your approach is highly applicable. I can you give you a mention on the (GT40) forum if you would like.
I used some cheap 350z front wheel bearings on a custom front suspension I built a few years ago, they lasted around 300 miles before they developed significant play. I replaced them witn SKF units. I took the easier option of basically copying MX5 front geometry & fabricated the uprights + wishbones to suit.
I was surprised by the differences. The cheap hubs had thinner wheel mount faces and while they're dimensionally the same, they feel different. They're looser than the SKF's. Did you weld your upright or have them machined?
I applaud you for tackling this complicated and slippery challenge, and I hope it works out very well for you. However, the C4 Corvette was known as an excellent handling car, so I'd be careful about thinking that you've discovered that its front suspension geometry is garbage. It was particularly good on the race track, where the 'problems' you describe should be at their worst. Having said that, Corvette Engineering did redesign the C4's front suspension slightly for the 1988-96 cars, which got even wider tires than the 1984-87 cars had. Lotus Engineering consulted on the project. They widened the front suspension (the distance between the wheel hub faces) a couple of inches - about an inch on each side - by lengthening the A-arms. That necessitated a series of other adjustments, as you can now well imagine. The wider the tire, the less tolerant it is of camber change during suspension movement, and I suspect that's why this was done. Longer arms generally result in reduced camber change during suspension motion. They also, as I recall, _did_ reduce the scrub radius in that redesign, and I have a vague recollection that this was specifically suggested by the Lotus engineers. Scrub radius is part of the 'black art' of tuning an overall front suspension design package, and tires had come an _awfully_ long way since the last time the Corvette guys had redesigned theirs, way back when they designed the 1963 C2 Corvette. So maybe their first C4 iteration wasn't absolutely perfect, but it was damned good. It definitely _wasn't_ all wrong, or junk, or stupid, or anything like that. Just saying...
Sorry, I was trying to say what you've said here. It's not garbage, it's fine for a C4 Corvette. The overall dynamics are meant to work in a system with the rear suspension. You're correct, they did change a few things in the later cars and the C5 and C6 are evolutions of this concept. I tried to stop short of saying it's bad - it's just suited to the car it was designed to work in and hacking away at it without doing the engineering is not a good idea. It also wasn't my idea - but the person that sold me the shortened cross member. I'm a little annoyed that they didn't take the time to work out the kinks so that it would work. I'm only a few thousand dollars in the hole because I didn't want to do this.
Wow, I love your honesty and the way you explained your process. From my experience, I would say most will run in loops because of the lack of tyre data available, as most things are fundamentally linked to those characteristics. It is interesting to use existing vehicles to determine your foundation numbers; however, it would say you should make sure you have similar spec tyres to those vehicles and have similar weight and centre of gravity. So, the critical pieces of data characteristics you are missing are the following Slip angle Vertical load Horizontal load Alignment torque Rolling resistance Another factor would be the aero characteristics of the vehicle. I look forward to seeing more videos from you. When I was a student, my go-to was Race Vehicle Dynamics. I think everything I've stated is within that book. Essentially, they say that the suspension is designed for the tyres and keeping them in their ideal window
Thanks - I hold fast to the notion that honesty is the only policy. We learn by our mistakes. In the next episode, I will expand a bit and try to resolve the selected variables from first principles. I was missing a few details that I've now got filled in and will work that through. It would have been too much in the first video!
Overall, this video is more in depth than than most suspension videos on TH-cam. And for people who want to learn more, books are the best option. The book i referenced most is from Derek Seward when building my engineering college race car suspension. As mentioned in other comments, it is important to look at the front and rear suspension as a whole. This is how you control the under/oversteer. Ultimately, if you are looking for the most performance out of your tires, then you need good tire data. Unfortunately tire companies are usually reluctant to give out this information. And testing the tires yourself is a lot of work. But if you have $500 burning a hole in your wallet, then you can get tire data from Tire test Consortium. But if you are trying to save money, lowering the upper a-arm chassis mount point about an inch should fix the camber problem.
Thanks - looking up the Seward book now. I'm not looking to fix the C4 - just trying to reset and do this myself. Tire data is the most difficult to get!
For sure - I have even walked away, thinking I have understood the chapter, then returned to some detail that just doesn't sound right. Like I don't like the M&M description of side view...it seems too simple. I understand the intersection of planes but the description of the motion can't be right. I've seen it - it's not a simple path - it's complicated. It probably also doesn't matter - but then again, maybe it does.
Great video and confirms what I had thought of the Corvette suspension design. It is surprising how many so called Sports Cars have understeer built into their suspension designs. When I was mocking up my Lotus chassis, I used PVC pipes to physically build the control arms that were cheep and super easy to modify. I found it helpful to see a physical model after reading books such as: (reading list sample for those interested in more) Race Car Vehicle Dynamics - Milliken & Milliken An Introduction to Race Car Engineering - Rowley Tune to Win - Smith Race Car Aerodynamics: Designing for Speed - Katz Automotive Aerodynamics - Howard Tyre and Vehicle Dynamics - Pacejka
Thanks - I've not read the last two - but the rest have been scoured. Understeer is "safe" or so I've been told. I was thinking along the same lines for a mock up but would use SendCutSend.com to sort out a wood model.
@@ThrottleStopGarage yea, I've been told understeer is safer, I can see the argument for a powerful car with novice drivers. But IMHO, it should should be able to adjust out the understeer. One of the many reasons I like the Mazda Miata, it has a solid platform that is widely adjustable at a reasonable cost. Really like what you're doing with your project, hope you have plans to take it to track days.
Have been following along since you started this project. FYI, those moog ball joints may be a bit stiff. After going through the task of redesigning my 66 mustang suspension, I discovered the ball joints were binding causing all of my driving tests and the conclusions I came to, were invalid. Glad to see you choosing a low friction joint. Nothing like having to start all over, again.
What a fantastic explanation for how to establish a starting point on geometry! If i remember correctly, your steering inner pivot should now fall on the line drawn between the inner pivot points of the upper and lower control arm.
Yes the length of the steering arms is correctly displayed in the last diagram. It should I theory fit in that line. But after determining this length, it will be offset by cause of Ackerman.
Just discovered this channel and this is truly amazing and fascinating. I'm an engineer who first started before my degree, and had to ask what your background is because I am genuinely impressed at your ability to ally practical manipulations and theory.
Most of its above my head and I'll likely never put this info into action, but count me in for the ride. I'm always keen to learn and build on existing basic knowledge of all types of vehicle design.
Fantastic stuff...really impressive. I've got a similar project ('66 122 wagon) I'm just starting to sketch out for an engine swap. This suspension piece is one of the things holding me up.
Thanks so much for sharing that hard earned information with us, sir. I'm just at the same stage with my build and wasn't confident with the direction I was going in. I'm utilising MX5 front and rear hubs on a scratch built Lotus 11 I,m currently building and was hoping use the same upper and lower control arms but with a different track width. I think I need to go downtrend route you've taken to haveany success...
I'm glad you're using A arm, wishbones. If you go strut, like that Nissan Z, you open another can of worms. OEM front ends seem to favour parking, and understeer, some even have different wheelbase lengths on each side of the car.
It was interesting to see what the scrub was and why they'd do that. I'm learning how many compromises exist in many of these cars. More than I would have thought.
I love your channel. I've wanted to build a car forever and this way I can build it vicariously through you. (I have the "Engineering To Win" book too.) One other thing: Considering the care and detail you go through I don't understand why you only have 46k subscribers. Rebuilding cars is fine, but honestly wouldn't hot rodders rather build their cars from just a donor body on up?
Excellent work, I'm going to get stuck in on mine once I get my body mounts done (thanks for the idea on casting your own by the way) so I'm sure of the fender clearances. I have slightly less variables than you as I need to keep the lower control arm frame side point on mine but the upper (within reason) and definitely the wheel side are all available. Thanks for compiling the data on things like KPA and Scrub out of all of those cars as a starting spot. Unclear if I start clean sheet on the front or practice by pulling the Q45 dimensions on the rear. Also why the 350Z hub for 5x4.5in [5x112.3mm]? I get not going with Ford hubs with the big centerbores but Nissan/Mazda/Hyundai are all still bigger than the Toyota/Honda size bore you have on your rear hubs. Keep up the great content!!!
Thanks - I still have to work out the side view - reading more textbooks right now. Stalling progress. Pull the rear dimensions and get a feel for it before tackling the front. That's basically what I've ended up doing. I also got lucky on the rear when I shortened it - no adverse effects. The 350Z hub is compact - about as compact as I've found and was on the same bolt circle (5 X 4.5") - it was also a rear wheel drive car - so the front hub doesn't have a drive spline. The difference is the hub center can be accommodated with a centering ring if needed.
I would quote Einstein I believe, "If you can' texplain it simply then you don't understand it well enough. "I think you did a smashing job of explanation! Well done!
Here you go: There are three books that I have read that I consider essential. Race Car Vehicle Dynamics by Milliken and Milliken www.sae.org/publications/books/content/r-146/ An Introduction to Race Car Engineering by Warren J. Rowley store.drpperformance.com/book-an-introduction-to-race-car-engineering-by-warren-j-rowley/ Fundamentals of Vehicle Dynamics by Gillespie www.sae.org/publications/books/content/r-114/?src=r-146 Read everything by Carroll Smith www.sae.org/publications/books/content/b-706.set/?src=r-146
I'm assuming you've seen the G-Comp Unser corvette upright as used on the StanceWorks Ferrari. These appear to be a good design that can be self fabricated.
Craig, where did you get the drawings for the Howe products? I need to use them on my suspension design and have not found stuff specifically for the Howes, only generalized 'K' number drawings on the pin height not locating shoulders and such.
Thanks to share your experience. I build small electric vehicle (1 seat and 2 seal) and my biggest problem is the suspension geometry. I rebuild 4 time the geometry for the 1 seat and now I designing for the 2 seat. I using negative scrub to plan to have a better driving experience at high speed. But you use positive Scrub is it better Positive Scrub? If you want, you can contact me and I will share my plans and feed back on the roads.
Negative scrub (from what I've read) is used on front wheel drive vehicles and those with split mu (diagonal) braking. RWD often has positive scrub. From what I've read (little experience) it's desirable to have scrub as part of the design. I'm saying that for me to get a handle on the clean sheet design, I needed to start somewhere. I found scrub definition was the right place for me to start.
One _more_ book to read: _How to Make Your Car Handle_ by Fred Puhn. (I didn't see it in the montage.) Might be unnecessary if you've read and understood all those others, but for viewers who are just getting started it should be a valuable tome.
Thinking the steering rack pivot width and steering arm pivot location should have been included in the discussion at this stage, for it’s coincidence with inner swing arm pivot points and steering rack location limitations. Great stuff and thanks for sharing.
I may not have explained that very well - that was the point of starting with the rack. Once I have those pivots, they must lie on the line intersecting the inner UCA/LCA line. So I started with the rack. I'll get this explained better after I complete the side view and work in the steering.
@@ThrottleStopGarage yes, depending on if the rack is front or rear of the hub of course. But the length and angle of the steering links must stay the same otherwise you will induce bump steer, but I feel like you know of this already. It felt as if you chose the length of the rack by placing 615mm between the inner upper en lower joints and that's that, but I surely hope to see more next video. :)
I see what you're saying - all I was trying to say was that you need to design around something. So I selected a narrow rack with a reasonable steering ratio that is available and used that as a start point. The pivot to pivot for the rack did control the LCA pivot placement as the rack needs to sit on the line described by the UCA and LCA inner pivot points. It needs to do this in 3D - it's currently just sitting there in 2D and is placed to eliminate bump steer. The Ackerman is then set in the top view design. There are a lot of things to consider!
HI, you said there was a lot of reading. Would you mind pointing me at a couple of these - names of books, where you got them, links to internet sources, anything you're happy to share. What you are doing is fantastic, keep pushing forward.
Thanks - There are three books that I have read that I consider essential. Race Car Vehicle Dynamics by Milliken and Milliken www.sae.org/publications/books/content/r-146/ An Introduction to Race Car Engineering by Warren J. Rowley store.drpperformance.com/book-an-introduction-to-race-car-engineering-by-warren-j-rowley/ Fundamentals of Vehicle Dynamics by Gillespie www.sae.org/publications/books/content/r-114/?src=r-146 Read everything by Carroll Smith www.sae.org/publications/books/content/b-706.set/?src=r-146
Awesome video, I see you have got .6deg per inch of camber and codos for figuring out this base line number. I believe your caster will also increase the camber as you turn the steering wheel, do you have a figure for angle of caster and how much extra this will add to the six degree per inch. Or is that in the next video, cant wait!
Thanks - great question. Yes - it's pretty weird what happened for me in doing this and I can't work it into a video. I've read these suspension and handling books for decades and I could tell anyone who would listen what the words all mean but (and this is a big but) my ability to visualize this as a coherent system was limited. The viewer that looked at the suspension on the jig and said it was nonsense, then pointed me to a video showing the car going around a corner to prove his point - he's got a very specific knowledge and talent I don't have. To get this to gel, I had to STOP reading the words and do the work to learn. I would explain this as the difference between knowing "that" something happens and "how/why" it happens. It hit me like a 1000 W bulb a few weeks ago and bang - I get it. Then it was like re-learning everything as I could make it work in my head. When we do the side view - we'll look at how caster adds/subtracts from camber in turns and the anti elements of the side view geometry (which are similar to the front view). There is a rate here as well - and you can get it in two different ways. I hope I can get it nailed down enough to explain it passably well.
Great video. Just my thoughts, but I'd try to design in adjustability. With those balljoints, you can get different lengths to tweak roll center, so that's good. Being able to adjust the inner pivot is good too as it will allow adjusting anti-dive and front/rear bias. Steering rack height adjustment to tweak bumpsteer. Alternative mounting holes or spacer stacks can be used to do it. Like the use of the Hub assemblies. I've been working on a design as and doing an IRS with it. Trying to find ones with the right wheel bolt patterns, good bearing spacing and readily available can be a hassle. Wonder if anybody has dug in deep to find some good ones?
I'm adding adjustability for sure. I hope to get it close, but I know it will need to be tuned. Finding the right bearing is challenging. The 350Z is pretty much perfect for my application.
I have experience.... messing suspension up real bad. Hahaha, I'm so curious to see the interplay between the camber gain and the aft kingpin inclination/offset planes of articulation in the SLA unit.
Very worth the wait! I have gone down part of this rabbit hole (the analysis side) and have noticed that factory suspensions are rarely what I expect. I really don't think much of the "millions spent" on car designs have very much to do with actual good performing suspension. I suspect ride comfort and stability/ease of use at low speeds trump actual road holding performance under rougher heavy usage... My question is, are you going to make full custom control arms, the send-cut-send way, or are you going to go looking for arms that exist that meet the specs you define, you know, mix n match from different makes? And have you planned to make the whole thing adjustable, with either shims of adjustable arms etc?
Thanks fellow rabbit hole adventurer. Full custom arms and uprights (may have these machined from billet - I'm unsure). With the wear parts (ball joints etc) being common enough parts that the rest should be fine. I'd like to add adjustability as otherwise I will get trapped. I have few delusions about being able to nail this.
Honestly, suspension geometry has always been a mystery to me and not something I've ever felt I could tackle. Your video has finally made some things click in my head.... It's also made me realise I do not want to design my own suspension! 😂 Do you think you would publish the designs when you finish?
In your research of existing cars roll centers to help decide your geometry, did you also see a pattern between roll center height and track widths? If your reference cars actually followed a 'rule'? You did not directly mention anything about that here. The closest is the camber curve which, of I remember correctly, helps to keep the COG over the roll center. I know it is impossible to cover everything in this length of video but curious as to if that was part of your process. Your classic car is narrower and taller than modern high performance cars so how much should your reference them? This probably could be a video in itself.
Interesting question. Track widths are wider these days and somewhat variable between car classes. The RCH has to be set in context to be sure and a value between 40 to 80 mm was pretty common. There are differences in how RCH is measured (force-based vs. geometric) and most of the forum material posted is from guys in their garage trying to sort things out (like me - and subject to uncertainty). What I found was they fell in that range. There is also variations in centre of gravity to consider. I'm unsure how closely I should follow - but there is variation in heights and there isn't a single value - same with camber curves. The Volvo had almost no camber gain or loss in it's original form - so guessing at a roll angle and then setting a camber curve is guess work. I had to pick one - I have raced cars with a lot of roll and it's not confidence inspiring - even on the road it's not what I prefer. I picked this rate based on this very limited experience.
I honestly thought it would be better than it is. I looked around and I could buy a cross member that I thought would fit and I was told by the company that this would be a nice solution. I can chalk this up to not wanting to do this. I thought I could buy something off the shelf and now I'm paying that price.
Are you sure your evaluation that test of the stock camber curve is correct ? At about 7:30 on. It looks to me like the wheel is going more neg camber with bump. Just eyeballing it compared to the background. Also your digital level has a generated image of a bubble level on it ? Its going to the right also, indicating the level is leaning to center of car. Or maybe im dyslexic .
I just went and checked it again and the stupid gauge was hung up on the lug nut. Dammit. I moved it a little and now the bubble is going the right way for the initial reading. The model in the computer is correct - the suspension in roll is losing contact patch. So what I said was right - what is shown is a bit of a mess.
@@ThrottleStopGarage Got ya ! So heres a question I have, regarding inputting suspension pivot pints into a computer model. In a situation like this, where the UCA pivot axis is inclined a fair bit downward (Anti dive), what point along that axis do you use for your 2D simulation ? The front ? The back ? its quite a difference in height above the ground plane. Love your videos. Thanks for the work. I am a suspension tinkerer also.
Thank you for explaining where some of the points in space come from and what is acceptable from a design point of view. From the way your thinking is progressing I am assuming you are going to machine your own uprights. From my experience (formula car racing) and research the suspension design starts with the upright, especially from the home workshop environment, as it is expensive to manufacture. It is usually more efficient to start with a production upright that has a minimum of compromises. Have you got one in mind? I am currently have the some of the same issues on my current project, a 77 Monza, which have crappy front suspension geometry, crappy uprights that don’t allow big brakes, good hubs, etc.
That's basically what I'm doing by defining the scrub and SAI first. These are upright variables. I'll probably get the uprights machined. Next up, side view...then I can sort that out.
I was having difficulty understanding why you picked the ball joint locations, but after some thought i believe the lower inner was dictated by where the rack would be and the uppers chosen for the camber curve desired. As an aside your basic geometry is almost the same as Carrol Smith shows in Engineer to Win.
@@randycollett1746 That's pretty much the case. I tried to locate the lower so that I would have clearance to the brake rotor (this was a challenge) and had it at around the same height as the LCA mount on the frame. These took a few tries to get right. I've read all Carrol Smith's books - while I didn't follow his method in this video - it's more or less Ch 17 in RCVD, I'm not surprised that they're similar. Things for sure gelled for me and a lot of this stuff started to make sense very recently.
Great video Craig. I'm wondering if something like a modernized Mustang II/Pinto would have worked. Seems to be pretty popular with hot rodders. Not sure if the geometry is there or not but it looks decent at a glance. Looking forward to the next vid 👍
@@ThrottleStopGarage If you have the skill and resources like you do a clean sheet bespoke design is the best answer for sure. Plus the cool factor is way higher :)
Are you still going to use the C4 vette spindle/upright or is that going to be made custom as well? I had planned on using the same setup in a 69 triumph GT6 but now I’m thinking it may not work since I need to narrow it so much. Great job on explaining your thoughts.
@@ThrottleStopGarage it seems very involved. I’m working on a triumph GT6 and trying to adapt a C4 suspension setup so am really enjoying the build. Thank you.
Another excellent video Craig. Are you trying to design a suspension that won't need adjusting, except for toe. Seems to me once you start adding shims or rotating cams all these parameters will change. Minutely, but will it be enough to blow your mind?
I think there has to be adjustability Alan - otherwise I'm going to get in trouble. I don't imagine that it's going to be right without the ability to move it around a bit.
@@ThrottleStopGarage Of course, I was just pulling your chain, didn't want your head to get too big, haha! Like everyone else here I'm in awe of your process. It's gotta be right or it's getting done again until it is. I saw a picture of you when I looked up the word patience in the dictionary, there you were!
No need to be looking at values from other cars except for sanity checks. Ask yourself what 'roll center height' actually means. You're essentially designing an 'anti-roll' feature; and that's a percentage, not some best practice value in millimeter or freedom parsecs. You are determining how much leverage you want to give the mass attached to the CG. Roll center on the tarmac means 0% anti-roll and a fun boat ride, RC at CG height means 100% anti-roll and zero roll even in a 5G turn. So you want to figure out the necessary amount of geometrically build-in 'anti-roll' to get x degrees of roll at x Gs of max lateral acceleration. The roll gradient for a boring family car is around 5°/G, a production sports car should be around 3-4°/G and little FSAE race cars are set up somewhere between 0.5 and 1.2°/G. My suggestion would be a target of 1.3G and 3°/G, pretty sporty for a road car. So the starting point for simulations I'd pick would be 4° of max roll, 0.25-0.5° static camber and 4.25-4.5° of camber gain at max roll. It's not a GT3 car after all. For the rest you'll need weight, weight distribution and spring rates. Get your spring rates sorted, calc the respective instant center heights, jacking forces, and see how much the car jacks up and rolls vs. what you want it to do, then apply the necessary 'anti-roll' by raising the static roll center to the right percentage of the CG height. CPI btw has a lot of interplay with caster, it basically increases (neg.) camber on the outside wheel and reduces it on the inside wheel. The right angle again depends on the amount of jacking, roll and positive camber gain at full droop. The good thing is, you can leave this parameter for last since you can simply slide the upper control arm along its virtual construction line once everything else is locked in. Hope this isn't too confusing, English isn't my first language.
Thanks for taking the time to write this out. I've gone over it several times - you've shared some very important thoughts. The RCH part is completely correct - my limitation is that I wasn't sure what that value should be for the percentage anti feature. The car originally had a RCH at ground level due to parallel (or close to it) control arm geometries and it was no fun. The method of "fixing" the suspension was to attach monster anti roll bars (25 mm - short arms) and increase the front spring rate. It was like a gocart. It lacked mechanical grip and I hated it. What I've been missing to do the calculations is a sense of what a reasonable G force would be - even the other parameters are more me guessing and I hate that. This is what is holding up the side view work right now - how much anti-dive is reasonable? The caster is the same issue - there is a set value (sanity check if you like) that seems to pair with a SAI. So for a SAI of 12° - I see a caster of around 5° where for less SAI, I've observed more caster. I need to compute that rate. Any advice on the side view construction parameters would be great. I understand that this is just "roll" in the pitch direction and that 100% would be alarming to drive as it would instantly react without any signaling to the driver. I think I've seen less than 20% tossed about as a start point but I have no idea if that's a good idea.
@@ThrottleStopGarage From my viewpoint you were gaining negative camber as the suspension was compressed....Adding spacers to the top arm imho will increase static positive camber.
This is just because of the camera perspective angle not being perfectly horizontal guys... You can clearly see at ride height it had 0.4 degree negative camber and at 1 inch travel it shifted the other way to 0.5 degree positive camber, which is the wrong direction. It should have been 1.5 degree negative for example
This threw me off a bit too as visually it looked like it was gaining negative camber. The arm lengths and angles tell the story though. The bottom arm is past center of travel and will lose effective length while the upper are is rotating into a longer effective length.
Hi Craig: You can check out my approach anytime. Get a bunch of random suspension parts from China, throw them in by eyeball and hope for the best. Not a lot of calculation since we are measuring to thin air anyhow. The electric car driveline has arrived.
Have you watched "Fanatic Builds"? I think he ended up designing his suspension from scratch. Quite involved and interesting in any case. P.S. Love your vids.
It seems like it's going in a better direction, the corvette suspension looked wierd to me as it had the upper control arm sloping outwards rather than inwards as is custom, therefore the corvette instant centre would've been outside the wheel. I wonder what the result of mounting the inside of the corvette upper control arm lower would've been, this would to my mind put the instant centre in a more conventional spot (therefore a better roll centre) and also give better negative camber gain. When you rolled the vehicle in the model it was still losing contact patch, perhaps it could use some more negative camber gain? Where is the steering rack in this? If you don't consider the location of the rack you'll likely get bumpsteer or rollsteer issues (you measured the width so I assume it was in the modeling). Rather than inputing standard figures for camber gain why not calculate what camber you need? Camber gain required is equal to the degrees of roll, you've got the track width and a reasonable idea of suspension travel which should be enough to calculate it with some trigonometry. An older euro vehicle will tend to be less track width so may not be the same as more modern. Also camber curves depend on which suspension it is, macpherson struts go entirely the wrong way and are forced to start with too much negative camber so they have some left in bump.
Yes - the IC's are outside the wheel. Things are not quite as I've shown them as the level was not flat on the hub and eagle-eyed viewers picked up what I was too tired to notice. The computer model is correct, it's a little not what you'd expect - point made. When I optimized it before, the 3D software had me moving a lot of things to get the attributes I wanted and it looked quite different. That's why I bought the stock crossmember - to find out what GM thought it should be. Steering rack is in place in the 2D model and designed around zero bump steer - probably the most important parameter. I'm just learning - the trick for me was not knowing what "reasonable" is with the parameter. So what's a reasonable lateral G? What sort of roll is not crazy? How little is too little? I'm all ears for feedback on these parameters. They are what is needed to do the entire thing properly. I am going to do this for the 3D model - but the 3D model needs inputs and I just needed to start. Does this make sense?
@@ThrottleStopGarage a lot of even amateur racers have lateral G's on thier videos, most seem to see less than 1.5G from memory. When I calculated roll for my suspension I used available travel in bump and droop from ride height, ie if you've 4" inches in bump and 4" inches in droop that's 8" across the track width of the car trigonometry means you can work out an angle of roll that would be the upper limit.
Although automotive companies spend millions on their designs, they have a lot more constraints than you do. You're building one custom vehicle for your own consumption, not trying to sell millions of vehicles to the masses. I realize you don't want to use an existing design, however you have the ability to leverage the knowledge others spent creating existing designs. This allow you to choose the benefits and compromises you're willing to accept based on your own criteria, parts availability, and packaging. I hope you see success on the suspension design and you're able to build something to surpass your expectations.
Couple things:
Roll center height alone is not a metric, it's in relation to the CG height. Yes spings and anti-swaybars control roll rate as well but then you get into ride quality. Also, roll centers need to be looked at in pair with the rear, creating a roll axis. This will tell you where weight is transferred. If the front RC is high but the rear is low, you'd get weight transfer on the outside rear, potentially lifting inside front and vice versa.
Most auto manufacturers design cars to understeer to get drivers to slow down. This can be done with roll rates (springs, swaybars), instant centers, or camber rates. So take those dimensions with a grain of salt and look more at the apsects of handling these cars have with what your going for and how the weight is balanced.
Before you finalize control arm and mount geometry, you will need to figure out where to put the steering rack. You are limited by its mounting height, but more importantly its length, and you may have to compromise on camber gain or RC height to minimize bump steer.
Great points. I'm not sure how to put this all into one coherent video. I'll do a little better with the RCH description as I know I oversimplified that point. And it's an important point. I did design with the steering rack in the loop - so from the front view, it's installed in the model and there is a bump steer output based on the FV only. It get's left alone until I do the top view then it's set. I have the bump steer dialed out at this time and when I put it into 3D then that parameter is (and this will start an argument) is probably the most important.
@@ThrottleStopGarage I did forget about being concise for the video.
Bump steer is definitely important and I'd focus on it more than camber gain. Going forward, I'd say also say bump steer is more critical than ackerman. You have slip angle regardless (just how tires work) and that slip angle changes depending on sidewall, tire pressure, intial camber, and tire load in the corner (so springs and swaybars). You dont have to be super precise with a numeber, but there is a range. Iirc the c5 corvette is ~130-140%. As a reference, my awd s10 (lowered and low pro tires) has the factory geometry of around 180% and the 320tw tires fight each other in parking lots.
Roll center axis does change lateral load weight transfer distribution yes, but does not change diagonally. Meaning, in the case of high rear roll stiffness, the outside rear tire may increase it's load but the front and rear weight distribution remains the same. Both the front tires loads could remain the same, but the inside rear decrease as much as the outside rear increases. To have longitudinal weight transfer, there must be longitudinal acceleration.
It's a common misinterpretation for people getting into vehicle dynamics
Also, roll centers are fake, but that's a deeper dive 😅
@sevenducecut the roll centre debate does my head in...I understand the desire to get a reference location, but the nature of the beast can't really be what is described. If you have specific advice...I'm all ears.
@ThrottleStopGarage You did your research and did just as well as anyone getting into this black magic.
Roll centers are a simplification of forces going through the control arms. The calculations of roll stiffness using roll centers are useful when you have a suspension system already made, and need to see what the system does as you make adjustments (on a car with adjustable points). A more complicated but closer to truth method uses instant center position to find jacking force, and IMO is better when first designing a system from scratch. I haven't seen this method in books, so I've had it on a video to do list for months 😅
I have watched you working this project for a while now, and several reiterations on the front suspension. Your explanation is actually making sense. Thanks
Glad to hear it! This one was particularly hard as I didn't find it intuitive to understand the process.
@@ThrottleStopGarage my step-father had a '64 AC Cobra with the leaf spring front suspension, I was going through the suspension while you were explaining it.
@@bradmaas6875 Cool - I wonder what the properties of that suspension are.
@@ThrottleStopGarage The front suspension was very similar to the AC Ace of the '50s, transverse spring acting as the upper control arms. Very long UCA's with the inner pivot functionally mobile during its travel. Might be time for some plywood mock-ups.
When I was in service, doing hundreds of maintenance and crash alignments over my 35 years, I kept an RC car with adjustable A-Arms, to demonstrate what you're showing. It had tiny bubble levels glued all over it. Great visual tool.
Cool - I've got a little model I made to prove the Ackerman relationship to steering rack position. I just couldn't visualize the complex relationship.
@@ThrottleStopGarage Are your "adjustment range" calculations accounting for vehicle loaded weight, driver and full gas? I'd be bummed if I ran out of caster or camber(ride height dependant). And trying to nail the spring rates will be added rocket science. Generally speaking, alignment spec loads are with 3/4 fuel and a 70kg driver, which I guess is out of date now. eyeroll.
@@kathysarmcandy1992 That's the idea...I've got several spreadsheets running with spring/roll and other calculations. I'm not sure how ultimately useful these are as they're copied formulae from others (or were donated by friends who have developed them). My guess is that they're close enough for a first approximation. It will need to be tuned when rolling.
@@ThrottleStopGarage Adjustable upper ball joints, I see similar to your screw-ins. Don't forget to make friends with an alignment shop. On and off hours after work and buy the shop pizza is totally worth it.
Xf motorsports did great videos on suspension design a few years back. The guys are super responsive over there too and helped me with my trans adapter plate design.
Ps they are also canadian
They've done some nice work.
Really appreciate the detailed breakdown of suspension dynamics. I've run into the same problem while trying to learn this; every resource I found was far too basic. The order of operations when designing a suspension from scratch is a game-changer. Thanks!
Glad it was helpful!
14-15 year old me (a third of a century ago, now...) designed a car, and I had that same wonky control arm setup like your C4 Corvette suspension on the bench.
My single source for guidance (from the public library) wasn't clear enough to me, and I tilted the top arm so the pivot on the top of the knuckle was lower than the pivot on the frame. Anyone roughing this out -- this was all pre-internet, so I was lucky to have a small town library with just that ONE book at the time.
Several years later, re-looking at my drawings in CAD, I mapped out the virtual center, and saw that camber curve: it was dished concave in, not convex out how we normally want for camber GAIN. That was my aha-moment to tilt the top arm the other way. We got stumped both by the camber curve.
Thanks for putting me back on track! Your explanation did clear up some of the murky less understood design process past where I stopped in my teens.
The visuals that Racing Aspirations provides helped backstop all the concepts better, as you explained. It probably helped that I had some, not all those concepts percolating in my head now for 30+ years after my first and only go at a suspension design, so as you explained it on video, I wasn't falling behind.
Very interested to see what the next step holds, how the 3d-analysis is done. Onto the voodoo and dark arts of anti-dive, anti-squat, and more! And, thank you again, for taking us along for the ride. :D
Thanks - it took a lot of work to get these concepts nailed into my head. I literally had to read and take notes and go out to the garage and play with things until how to design instead of analyse. I'm still not confident.
@@ThrottleStopGarage I get the slow grok on the concepts. I think you explained things well, especially all your emphasis on the things that were murky to understand from all the technical reading.
The only thing that seems unclear to me, just because of that oddball upper control arm geometry on it as it is right now:
Was the C4 setup like the one on the bench, really? with outward/downward sloping upper Control arms so the camber curve was all camber loss?
Or was this C4 setup modified by the seller in ways that made it set up so? :o
@chrstphrr the outward sloping UCA is how it was designed. That's how it is in the drawings and on the bench it is set so all points are located as they would be at ride height. There is another set of mount holes in the cross member that are lower. I've never seen anyone mount the dog bone down there, but it's tempting.
Well that certainly explains why I’ve always felt I was spinning round and round in successively smaller circles when I’ve tried to understand suspension design. Reading the books made my eyes glaze over. Sometimes after giving up I would even head for the donut shop for a superior glazed solution. Unfortunately that never helped either. You’ve inspired me to give it a go. Definitely looking forward to the next episode. In the meantime I’ll be rewatching your previous episodes and reading all the comments to help drive this home. I’ll probably even pull those dusty books back off the shelf for another spin. Thanks so much Craig for slogging through this. 🙏🙏🙏🙏
Thanks - this was a game changer for me. Like you, I could read the words and even explain it, but the shapes wouldn't work in my head. I had no "sense" of things like the viewer who pointed out the problem in a suspension by looking at it. Gold star for him. This did come to me like a clap of lightning. Boom - suddenly it all snapped into place.
This is the suspension video I've been hunting for years
Don't measure from the tip of a tape measure if you care for good accuracy, "burn an inch" or perhaps 10mm in your case and measure using the divisions and not the fairly approximate tip of the tape :)
The effort you're going into for this is pretty impressive and your willingness to jump into areas where you know you need to learn is great! I will admit the suspension stuff is Greek to me although I've certainly heard many of the terms and have some vague understanding. Much appreciation for trying to explain it as you're learning! The feedback you'rew getting is also awesome and I cannto wait to see what you end up building and how you like driving it eventually afterwards. It's like you're building Tally Ho in automative form - I love it!
Thanks - I'm lucky that I get the great feedback that I get. It all helps.
I know you already replied that the level was not placed as expected on the hub, but I think it’s important to notice that you were reading the “bubble” level backwards. As you increased bump on the C4 suspension mock up, camber was going more negative. I watched a promo video for the tool to confirm. I don’t mean that this is the right setup for your car, but I don’t want you to be confused and frustrated in the future when using that tool.
Also, you are embarking on a long journey if you plan to fabricate all the components. Mike Burrows from the channel Stance Works did something similar and re-did it a few times. He has some good videos to reference on the fab techniques and learnings.
Good luck!
Thanks - I noticed that after the video was released. I had moved the level to shoot the segment. It happens. I understand the nature of the journey - looking forward to most of it whatever happens.
I'm glad I found this video. I do want to do a swap as my car has a solid axle type and parts are becoming rarer. A custom suspension design would allow me to build it with easier to find parts.
Talking about camber gain around 9min. Actually looks like you went from a half degree postitive camber to negative 1.5 with bump (i assume your angle gauge 's bumb feature should be same as a level bubble). Regardless of that, its still not enough static camber at ride height or enough gain through the travel.
Such a low roll center is counter to where I'm going through with testing on my car. However yours is WAY lighter than mine. If you have the room don't be afraid to add chassis side adjustment for the lower arm that's higher up. Which will directly raise the roll center. Up and down adjustment for the upper arm is a good idea too. I was once told anything that can be made adjustable is a good move.
The camber gain target you showed seems too little if you plan spirited driving. Super basic but i set mine by sitting on the fender and using the angle cube to measure body roll (2.4), then set the cube on the tire to check for gain or loss. I went with a degree more negative gain than body roll as a place to start. Best year I've had yet in tire longevity/wear doing street, autocross, and track days. I don't know what that gain is vs compression travel but it is definitely more than .6/inch.
However a street only setup probably wint need the additional camber gain. In aggressive driving it does need more because you also have to compensate for tire deflection.
Overall, good video!
I also spent months researching just to start understanding what's all going on when everything moves.
When I filmed the level, it wasn't flat to the hub and I didn't notice. The computer model is correct and it's not great. I am still working on the roll and camber curve. This will get more attention as I do the side view.
@@ThrottleStopGarage
Oh, and if you haven't figured it out yet. The lower arm primarily locates the roll center height and the upper dictates camber gain. Yes, both affect on the other but to a much lesser effect.
@@ThrottleStopGaragesomeone else was commenting about the rack placement.
Easiest/obvious placement is with the tie rod pivots lined up with the lower control arm pivots. Then place the outer tie rod lined up with the ball joint. If they follow the same arc then it can't bump steer beyond some pretty minor differences.
@@codysmithmotorsports736 I'll go over that detail in the next video...I think.
Thank you Craig for doing what few have the patience for. Basics, camber gain; when you "bumped" the wheel up first 1" and then 3", you lost negative camber gain, you didn't gain camber. Maybe you just misspoke there but it's an important point when we're talking negative vs positive, losing vs gaining, concave vs convex bla bla. The upper control arm is obviously shorter than the lower, as both move through their respective arc ranges, the shorter arm has no choice but to result in offering "negative camber gain". The problem is the upper and lower arms are not parallel, causing some "loss" at different points through the range of travel, but ultimately will give negative camber gain. The geometry has no choice but to produce this result, the upper arm is shorter. In the Racing Aspirations software it is clearly set up differently to what you have on the bench in your garage, in the software version the arms are very nearly parallel, the Corvette is very clearly not parallel. Sorry if this sounds like I'm angry, really I'm not, I'm only wanting to clarify why you lost camber gain, the upper arm begins, lets call it drooped, the lower arm has some degrees of "rise" from the "at rest" starting positions of your experiment to see what the suspension was physically giving you. This is where the dark arts show up, bump steer, weight jacking, roll center, roll axis on and on, this is why this is all so difficult to master. Before you move on to the side view, you have more work to do from the front view. Ford went to an extreme to "correct" for roll center migration, they designed the infamous Twin Torsion bar suspension. This provided for a really smooth soft supple ride, but did nothing for handling or tire wear. It was a cheap solution or alternative to Chevy's double A arm suspension, which is as you are finding out, very complicated to mitigate all the negative compromises required for mass produced suspension. Remember the Corvair, lawsuits were abundant because the geometry/design was flawed from the beginning.
As important as reading and studying the available printed material is, like another "commenter" said, an RC, radio controlled toy car can teach you as much as all those books combined. I highly recommend getting a RC car to test your theories/setup/geometry.
Don't get me started on roll center, every race car mechanic has their own opinion of what is the optimal height and they have tested their theories on the track, go to a local circle track and talk to those guys, invaluable. Pavement or asphalt racers would be optimal as you want your car to run on the street. Oh, and potholes, don't forget our Alberta roads mostly suck.
Thanks - I did fumble with the words a bit there and the level was hung on the edge of the lug nut and showing the opposite of what is happening at ride height - that segment is a bit of a mess. Viewers caught it - I'm too tired. The computer model is correct and the path is funky for the C4 - good enough and I can sleep well enough.
4 am time to watch 40 minute video
I'm so sorry!
@@ThrottleStopGarage don't be. it was worth it. No matter when you upload somewhere will be 4 am.
Cool video! I use the C4 front suspension for my race car, and I did a few things to help (not solve) the issue that you observe. My first solution was to get a longer lower control arm ball joint from Howe. Does put the control arm close to wheel hoop though, but 18" wheels clear fine. I also chucked the upper control arm cross bar spacers in the lathe and cut them down to the absolute minimum in the back and enough in the front that I could run more caster than stock for camber gain under steering input. Still doesn't give more than bout 2 degrees maximum static camber and I am going to go to a custom shorter upper control arm to get that sorted. Lowering the upper arm inboard pickup points would help as well, but not sure how much fabrication I'm willing to undertake.
Nice video, thanks for making it!
Cool. I've got a set of those taller lower ball joints (1/2" if memory serves) that were part of my original optimization quest.
When I was autocrossing my C4 ages ago, I put offset lower control arm bushings in which increased the static negative camber markedly. These are still readily available.
FDF RaceShop in Ontario makes angle kits for drift cars (aka a lot of experience with non-OEM front suspension geometry modification). I believe they have their own engineering department too. Might be worth trying to reach out to them… they’d probably be atleast a little interested in helping out upon hearing you built a 60’s Volvo body out of carbon fiber
Great video. Really appreciate the visual representations! Probably a huge PIA to film and record
Thanks - there are a few companies around that could help out.
This is incredible. You did a great job of explaining! I feel like anyone who is doing the same will save TONS of time and money by watching. Remember, if the women don't find ya handsome, they can at least find ya handy.
Thanks! I wonder if duct tape will sponsor an episode?
Well done, easy enough to follow. Thanks for putting in the weeks so we can put in the hours!
Glad it was helpful!
I really need this. I am very reluctant to continue ‘trying’ to work with 40+ year old suspension parts and constraints, especially when it is well known that bump steer was an issue back then. I actually like the idea of starting from scratch on my mk7 Bolwell.
Glad it is helping.
When poking around with suspension camber gain / loss, I've used a peg board and furring strips to simulate different A arm locations and lengths.
I've done the same - the combo of the web based platform and the CAD software now replace much of this, but getting a "feel" is important for me.
Great video and a good attempt or start at solving your problems. I am not aware what tires you plan on running. If they have a larger side wall and will have significant contact patch migration under cornering load, then you may need to increase the scrub radius. You want to insure that the forces generated at the contact patch stay on the same side of your SIA as you load it up dynamically. You don’t want to have the forces at the contact patch go past the SIA, as when this happens any feedback your feeling in the steering wheel will also change direction, pulling the steering wheel in the opposite direction mid corner.
Next thing, as the SIA increases, you get more / less camber gain as you turn the steering wheel. So if your benchmark cars were getting 0.6° camber gain, but had a lower SIA angle, you might need to design in more or less camber gain in your wishbone length ratios, to accomodate for the camber loss you get when turning the wheel with a high SIA angle.
Looking forwards to the next video.
Cheers.
Ryan.
Thanks Ryan - 225 45 17's are the target tire - probably a Michelin Pilot Sport 4S or similar. Is there a source that I should read to take a stab at ensuring that the scrub is in the right range for the tire. I've never been able to squeeze tire data from the MFG's (and I've tried). I wonder if there is a simple tire model that would be good enough for a first approximation?
@@ThrottleStopGarage I have not looked. I know Avon motorsport did have some tire data available for their motorsport tires. You might also be able to get data from some of the other motorsport tire suppliers. I've never tried. I think as long as we are aware of it, and have some idea about where it will be, then ultimately, we can design for it. I'm going to assume you have seen some GoPro type footage somewhere that shows the tire being stretched, pulled across from its static position and heavily loaded mid corner? If we use that as our crude basis for estimation we can then say we assume that the tire side wall goes from being vertical/perpendicular to the surface, to say at absolute worse case, 60° from perpendicular. If we know the side wall height (difference between tire radius and rim radius) then we can calculate how the contact patch may migrate laterally under load. All I am concerned about is keeping this contact patch on the side of the SAI during cornering. There is also a side view component to this as well. Caster and hub offset forwards or rearwards of the SIA come in to play in this orientation. My thinking is that making sure the contact patch is in a known position relative to the SIA is still valid to keeping a consistent feel in the steering under all conditions. If you go back to the C4 corvette. The tire size was 255x45 R17. So a side wall height of 255*0.45 = 114.75mm? (is it radius or diameter?) * Sin 60° = ~99mm. IE under load the stock Vett front contact patch may move up to 100mm or 4" assuming the 114.75mm was the radius. Hence why its knuckle offset may have been where it was.
The explanation of front view design is great, start with the track width you need, add a few locations at the wheel, start from the center of the car then connect the dots.
I'll be waiting for the side view and am hoping for a top view vid as well. Concerning top view, most if not all street cars ( use a 77 to 90 full sized GM ) have the lower control arm rotated so the front pivots are closer together. I'm not sure why this is done but I think it interacts with what I'm typing below. The upper is rotated as well but I'd have to look at a suspension to see what way.
As for side view, the control arms pivots are tilted relative to each other , I'm pretty sure this has to do with anti dive where braking forces try to add spring rate to the front suspension. Think of it this way. If the brake was locked and you applied a torque in the direction of forward motion , how much torque would it take to move the suspension into droop? High torque = minimal anti dive = nose droops a lot under braking. Low torque = high anti dive = minimal nose droop = possibly too high of an effective spring rate leading to turn in problems.
Yup - working on these two views now. I'm going to start with a review of the roll and pitch centre to clear up that point.
Thanks Throttle Stop Garage ... very, very well thought through and articulated ... about the most helpful suspension geometry video I have seen. As I am building a GT40 your approach is highly applicable. I can you give you a mention on the (GT40) forum if you would like.
I used some cheap 350z front wheel bearings on a custom front suspension I built a few years ago, they lasted around 300 miles before they developed significant play. I replaced them witn SKF units. I took the easier option of basically copying MX5 front geometry & fabricated the uprights + wishbones to suit.
I was surprised by the differences. The cheap hubs had thinner wheel mount faces and while they're dimensionally the same, they feel different. They're looser than the SKF's. Did you weld your upright or have them machined?
I applaud you for tackling this complicated and slippery challenge, and I hope it works out very well for you. However, the C4 Corvette was known as an excellent handling car, so I'd be careful about thinking that you've discovered that its front suspension geometry is garbage. It was particularly good on the race track, where the 'problems' you describe should be at their worst.
Having said that, Corvette Engineering did redesign the C4's front suspension slightly for the 1988-96 cars, which got even wider tires than the 1984-87 cars had. Lotus Engineering consulted on the project. They widened the front suspension (the distance between the wheel hub faces) a couple of inches - about an inch on each side - by lengthening the A-arms. That necessitated a series of other adjustments, as you can now well imagine. The wider the tire, the less tolerant it is of camber change during suspension movement, and I suspect that's why this was done. Longer arms generally result in reduced camber change during suspension motion.
They also, as I recall, _did_ reduce the scrub radius in that redesign, and I have a vague recollection that this was specifically suggested by the Lotus engineers. Scrub radius is part of the 'black art' of tuning an overall front suspension design package, and tires had come an _awfully_ long way since the last time the Corvette guys had redesigned theirs, way back when they designed the 1963 C2 Corvette.
So maybe their first C4 iteration wasn't absolutely perfect, but it was damned good. It definitely _wasn't_ all wrong, or junk, or stupid, or anything like that. Just saying...
Sorry, I was trying to say what you've said here. It's not garbage, it's fine for a C4 Corvette. The overall dynamics are meant to work in a system with the rear suspension. You're correct, they did change a few things in the later cars and the C5 and C6 are evolutions of this concept. I tried to stop short of saying it's bad - it's just suited to the car it was designed to work in and hacking away at it without doing the engineering is not a good idea. It also wasn't my idea - but the person that sold me the shortened cross member.
I'm a little annoyed that they didn't take the time to work out the kinks so that it would work. I'm only a few thousand dollars in the hole because I didn't want to do this.
Wow, I love your honesty and the way you explained your process. From my experience, I would say most will run in loops because of the lack of tyre data available, as most things are fundamentally linked to those characteristics. It is interesting to use existing vehicles to determine your foundation numbers; however, it would say you should make sure you have similar spec tyres to those vehicles and have similar weight and centre of gravity.
So, the critical pieces of data characteristics you are missing are the following
Slip angle
Vertical load
Horizontal load
Alignment torque
Rolling resistance
Another factor would be the aero characteristics of the vehicle.
I look forward to seeing more videos from you.
When I was a student, my go-to was Race Vehicle Dynamics. I think everything I've stated is within that book. Essentially, they say that the suspension is designed for the tyres and keeping them in their ideal window
Thanks - I hold fast to the notion that honesty is the only policy. We learn by our mistakes. In the next episode, I will expand a bit and try to resolve the selected variables from first principles. I was missing a few details that I've now got filled in and will work that through. It would have been too much in the first video!
I love this project! And especially the approach you take to every part of it
Thanks so much!
I’m only halfway through the video and I gotta say this is a tremendous amount of work. You’re right. I don’t want to do this. Thanks for doing this.
Much appreciated. It's been exhausting. I only have the very basics down and am still learning as fast as I can.
By far the most detailed instructional/educational on automotive suspension design video . Awesome!
Glad you liked it!
Overall, this video is more in depth than than most suspension videos on TH-cam. And for people who want to learn more, books are the best option. The book i referenced most is from Derek Seward when building my engineering college race car suspension.
As mentioned in other comments, it is important to look at the front and rear suspension as a whole. This is how you control the under/oversteer.
Ultimately, if you are looking for the most performance out of your tires, then you need good tire data. Unfortunately tire companies are usually reluctant to give out this information. And testing the tires yourself is a lot of work. But if you have $500 burning a hole in your wallet, then you can get tire data from Tire test Consortium.
But if you are trying to save money, lowering the upper a-arm chassis mount point about an inch should fix the camber problem.
Thanks - looking up the Seward book now. I'm not looking to fix the C4 - just trying to reset and do this myself. Tire data is the most difficult to get!
Nice job man… We are all counting on you.
I appreciate that!
I’ve found that I had to read many many books to get the whole picture. No one book by its self made these topics make sense to me.
For sure - I have even walked away, thinking I have understood the chapter, then returned to some detail that just doesn't sound right. Like I don't like the M&M description of side view...it seems too simple. I understand the intersection of planes but the description of the motion can't be right. I've seen it - it's not a simple path - it's complicated. It probably also doesn't matter - but then again, maybe it does.
Wow! I don't know if you've done this correctly, but it sure makes sense. Thanks for the heavy lifting. I'm dieing to see the Volvo on the track.
It's the best I can come up with...until the next video when I revise a few of these things and move to side view.
Great video and confirms what I had thought of the Corvette suspension design. It is surprising how many so called Sports Cars have understeer built into their suspension designs. When I was mocking up my Lotus chassis, I used PVC pipes to physically build the control arms that were cheep and super easy to modify. I found it helpful to see a physical model after reading books such as: (reading list sample for those interested in more)
Race Car Vehicle Dynamics - Milliken & Milliken
An Introduction to Race Car Engineering - Rowley
Tune to Win - Smith
Race Car Aerodynamics: Designing for Speed - Katz
Automotive Aerodynamics - Howard
Tyre and Vehicle Dynamics - Pacejka
Thanks - I've not read the last two - but the rest have been scoured. Understeer is "safe" or so I've been told. I was thinking along the same lines for a mock up but would use SendCutSend.com to sort out a wood model.
@@ThrottleStopGarage yea, I've been told understeer is safer, I can see the argument for a powerful car with novice drivers. But IMHO, it should should be able to adjust out the understeer. One of the many reasons I like the Mazda Miata, it has a solid platform that is widely adjustable at a reasonable cost. Really like what you're doing with your project, hope you have plans to take it to track days.
Have been following along since you started this project.
FYI, those moog ball joints may be a bit stiff. After going through the task of redesigning my 66 mustang suspension, I discovered the ball joints were binding causing all of my driving tests and the conclusions I came to, were invalid. Glad to see you choosing a low friction joint. Nothing like having to start all over, again.
I was surprised how stiff they were! I'm only using them for mock up and for making the arms.
We like rabbit holes…. Its nicer when it’s someone else’s because we can just move on where you are stuck at least for now 👏
LOL - I should name this car "Project Rabbit Hole"
What a fantastic explanation for how to establish a starting point on geometry!
If i remember correctly, your steering inner pivot should now fall on the line drawn between the inner pivot points of the upper and lower control arm.
Yes - that's correct. That is why the steering rack choice comes before setting the LCA/UCA inner pivot relationship. I didn't appreciate this before!
Yes the length of the steering arms is correctly displayed in the last diagram. It should I theory fit in that line. But after determining this length, it will be offset by cause of Ackerman.
Just discovered this channel and this is truly amazing and fascinating. I'm an engineer who first started before my degree, and had to ask what your background is because I am genuinely impressed at your ability to ally practical manipulations and theory.
I'm a scientist.
Most of its above my head and I'll likely never put this info into action, but count me in for the ride. I'm always keen to learn and build on existing basic knowledge of all types of vehicle design.
Glad to have you onboard!
Great work, well done. Look forward to the next 'teaching' session.
More to come! I have to crack the nut of the side and top view geometries. That should be next.
That was fun! Can't wait for the next installment!
Thanks - I have to figure out the side view now...going as fast as I can.
Eagerly waiting for new updates.
On it hard right now! Side view coming with revisions to the front view method.
Looks like a great start of v2.0
I sure hope so.
As always you did not skip any steps. I think that even I could understand it. 👍
Thanks - it's been quite the adventure and this learning was hard for me.
Fantastic stuff...really impressive. I've got a similar project ('66 122 wagon) I'm just starting to sketch out for an engine swap. This suspension piece is one of the things holding me up.
Love the wagon. What engine are you swapping?
A mercedes M113 5L. It's a tight fit, but if I can just figure out what to do about those upper control arms...she should shoehorn in!
Thanks so much for sharing that hard earned information with us, sir. I'm just at the same stage with my build and wasn't confident with the direction I was going in.
I'm utilising MX5 front and rear hubs on a scratch built Lotus 11 I,m currently building and was hoping use the same upper and lower control arms but with a different track width. I think I need to go downtrend route you've taken to haveany success...
I am glad this was helpful. I'm just working on a better method for determining roll and pitch centres. Next video will have these details.
Awesome work! Man, how in the hell do you get all that done in such a small space? Bravo!!!
Thank you very much! I'm used to the tiny space. I hope to get a larger garage built in the future.
He probably cusses a lot.
😁
I'm glad you're using A arm, wishbones. If you go strut, like that Nissan Z, you open another can of worms. OEM front ends seem to favour parking, and understeer, some even have different wheelbase lengths on each side of the car.
It was interesting to see what the scrub was and why they'd do that. I'm learning how many compromises exist in many of these cars. More than I would have thought.
I love your channel. I've wanted to build a car forever and this way I can build it vicariously through you. (I have the "Engineering To Win" book too.) One other thing: Considering the care and detail you go through I don't understand why you only have 46k subscribers. Rebuilding cars is fine, but honestly wouldn't hot rodders rather build their cars from just a donor body on up?
Thanks, much appreciated. I don't think I'll ever hit 100k subs. As long as we're all having fun, it's fine.
Excellent work, I'm going to get stuck in on mine once I get my body mounts done (thanks for the idea on casting your own by the way) so I'm sure of the fender clearances. I have slightly less variables than you as I need to keep the lower control arm frame side point on mine but the upper (within reason) and definitely the wheel side are all available. Thanks for compiling the data on things like KPA and Scrub out of all of those cars as a starting spot. Unclear if I start clean sheet on the front or practice by pulling the Q45 dimensions on the rear.
Also why the 350Z hub for 5x4.5in [5x112.3mm]? I get not going with Ford hubs with the big centerbores but Nissan/Mazda/Hyundai are all still bigger than the Toyota/Honda size bore you have on your rear hubs.
Keep up the great content!!!
Thanks - I still have to work out the side view - reading more textbooks right now. Stalling progress. Pull the rear dimensions and get a feel for it before tackling the front. That's basically what I've ended up doing. I also got lucky on the rear when I shortened it - no adverse effects.
The 350Z hub is compact - about as compact as I've found and was on the same bolt circle (5 X 4.5") - it was also a rear wheel drive car - so the front hub doesn't have a drive spline. The difference is the hub center can be accommodated with a centering ring if needed.
Great explanation looking forward to the next episode
Glad you enjoyed it
I would quote Einstein I believe, "If you can' texplain it simply then you don't understand it well enough. "I think you did a smashing job of explanation! Well done!
Thanks - aside from the digital level SNAFU - the rest is correct. There is a lot left to cover and I'm still learning at this point.
Excellent work Craig!
Thank you! Cheers!
Thanks for the details! I would love a litterature list!
Here you go:
There are three books that I have read that I consider essential.
Race Car Vehicle Dynamics by Milliken and Milliken
www.sae.org/publications/books/content/r-146/
An Introduction to Race Car Engineering by Warren J. Rowley
store.drpperformance.com/book-an-introduction-to-race-car-engineering-by-warren-j-rowley/
Fundamentals of Vehicle Dynamics by Gillespie
www.sae.org/publications/books/content/r-114/?src=r-146
Read everything by Carroll Smith
www.sae.org/publications/books/content/b-706.set/?src=r-146
@@ThrottleStopGarage awesome! :D
Great video, I'm looking forward to seeing your upright design using the 350Z hubs
You and me both! I'm still working on the side view and a more coherent treatment of the front view roll centre placement.
I'm assuming you've seen the G-Comp Unser corvette upright as used on the StanceWorks Ferrari. These appear to be a good design that can be self fabricated.
Craig, where did you get the drawings for the Howe products? I need to use them on my suspension design and have not found stuff specifically for the Howes, only generalized 'K' number drawings on the pin height not locating shoulders and such.
This is incredibly helpful.
I'm glad it helps.
Awesome work and an unbelievable amount of time! Well done!
Thank you very much!
Wow, just wow!
Thanks!
Thanks to share your experience. I build small electric vehicle (1 seat and 2 seal) and my biggest problem is the suspension geometry. I rebuild 4 time the geometry for the 1 seat and now I designing for the 2 seat. I using negative scrub to plan to have a better driving experience at high speed. But you use positive Scrub is it better Positive Scrub? If you want, you can contact me and I will share my plans and feed back on the roads.
Negative scrub (from what I've read) is used on front wheel drive vehicles and those with split mu (diagonal) braking. RWD often has positive scrub. From what I've read (little experience) it's desirable to have scrub as part of the design. I'm saying that for me to get a handle on the clean sheet design, I needed to start somewhere. I found scrub definition was the right place for me to start.
One _more_ book to read: _How to Make Your Car Handle_ by Fred Puhn. (I didn't see it in the montage.) Might be unnecessary if you've read and understood all those others, but for viewers who are just getting started it should be a valuable tome.
Thinking the steering rack pivot width and steering arm pivot location should have been included in the discussion at this stage, for it’s coincidence with inner swing arm pivot points and steering rack location limitations.
Great stuff and thanks for sharing.
I may not have explained that very well - that was the point of starting with the rack. Once I have those pivots, they must lie on the line intersecting the inner UCA/LCA line. So I started with the rack. I'll get this explained better after I complete the side view and work in the steering.
You say they must lay on the line intersecting UCA/LCA line, but if you do this you have 0 Ackerman, just so you know and don't oversee this.
You can get Ackerman with relative location of the rack to the centre line of the hub - the further out you go the more you get.
@@ThrottleStopGarage yes, depending on if the rack is front or rear of the hub of course. But the length and angle of the steering links must stay the same otherwise you will induce bump steer, but I feel like you know of this already.
It felt as if you chose the length of the rack by placing 615mm between the inner upper en lower joints and that's that, but I surely hope to see more next video. :)
I see what you're saying - all I was trying to say was that you need to design around something. So I selected a narrow rack with a reasonable steering ratio that is available and used that as a start point. The pivot to pivot for the rack did control the LCA pivot placement as the rack needs to sit on the line described by the UCA and LCA inner pivot points. It needs to do this in 3D - it's currently just sitting there in 2D and is placed to eliminate bump steer. The Ackerman is then set in the top view design. There are a lot of things to consider!
Great episode!
Thanks.
Excellent video.
Thank you very much!
HI, you said there was a lot of reading. Would you mind pointing me at a couple of these - names of books, where you got them, links to internet sources, anything you're happy to share. What you are doing is fantastic, keep pushing forward.
Thanks - There are three books that I have read that I consider essential.
Race Car Vehicle Dynamics by Milliken and Milliken
www.sae.org/publications/books/content/r-146/
An Introduction to Race Car Engineering by Warren J. Rowley
store.drpperformance.com/book-an-introduction-to-race-car-engineering-by-warren-j-rowley/
Fundamentals of Vehicle Dynamics by Gillespie
www.sae.org/publications/books/content/r-114/?src=r-146
Read everything by Carroll Smith
www.sae.org/publications/books/content/b-706.set/?src=r-146
Awesome video, I see you have got .6deg per inch of camber and codos for figuring out this base line number. I believe your caster will also increase the camber as you turn the steering wheel, do you have a figure for angle of caster and how much extra this will add to the six degree per inch. Or is that in the next video, cant wait!
Thanks - great question. Yes - it's pretty weird what happened for me in doing this and I can't work it into a video. I've read these suspension and handling books for decades and I could tell anyone who would listen what the words all mean but (and this is a big but) my ability to visualize this as a coherent system was limited. The viewer that looked at the suspension on the jig and said it was nonsense, then pointed me to a video showing the car going around a corner to prove his point - he's got a very specific knowledge and talent I don't have.
To get this to gel, I had to STOP reading the words and do the work to learn. I would explain this as the difference between knowing "that" something happens and "how/why" it happens. It hit me like a 1000 W bulb a few weeks ago and bang - I get it. Then it was like re-learning everything as I could make it work in my head.
When we do the side view - we'll look at how caster adds/subtracts from camber in turns and the anti elements of the side view geometry (which are similar to the front view). There is a rate here as well - and you can get it in two different ways. I hope I can get it nailed down enough to explain it passably well.
Appreciate your effort
Thanks a lot
Wow man... incredible!
Thanks a lot!
Great video. Just my thoughts, but I'd try to design in adjustability. With those balljoints, you can get different lengths to tweak roll center, so that's good. Being able to adjust the inner pivot is good too as it will allow adjusting anti-dive and front/rear bias. Steering rack height adjustment to tweak bumpsteer. Alternative mounting holes or spacer stacks can be used to do it.
Like the use of the Hub assemblies. I've been working on a design as and doing an IRS with it. Trying to find ones with the right wheel bolt patterns, good bearing spacing and readily available can be a hassle. Wonder if anybody has dug in deep to find some good ones?
I'm adding adjustability for sure. I hope to get it close, but I know it will need to be tuned. Finding the right bearing is challenging. The 350Z is pretty much perfect for my application.
I have experience.... messing suspension up real bad. Hahaha, I'm so curious to see the interplay between the camber gain and the aft kingpin inclination/offset planes of articulation in the SLA unit.
I'm also interested to see how this goes in side view and then all together.
I’m pulling for you; we’re all in this together
Thanks!
It's Red Green approved !
@@alanmony1582 I'm a man, but I can change...if I have to...I guess.
I would’ve gone with the Corvette suspension and just drove into a ditch. Nice work!
LOL - I know you're kidding.
Very worth the wait! I have gone down part of this rabbit hole (the analysis side) and have noticed that factory suspensions are rarely what I expect. I really don't think much of the "millions spent" on car designs have very much to do with actual good performing suspension. I suspect ride comfort and stability/ease of use at low speeds trump actual road holding performance under rougher heavy usage...
My question is, are you going to make full custom control arms, the send-cut-send way, or are you going to go looking for arms that exist that meet the specs you define, you know, mix n match from different makes? And have you planned to make the whole thing adjustable, with either shims of adjustable arms etc?
Thanks fellow rabbit hole adventurer. Full custom arms and uprights (may have these machined from billet - I'm unsure). With the wear parts (ball joints etc) being common enough parts that the rest should be fine. I'd like to add adjustability as otherwise I will get trapped. I have few delusions about being able to nail this.
Honestly, suspension geometry has always been a mystery to me and not something I've ever felt I could tackle. Your video has finally made some things click in my head.... It's also made me realise I do not want to design my own suspension! 😂 Do you think you would publish the designs when you finish?
I understand completely. I can share the design when done.
Excellent info here!! Keep it coming!!
Thanks! Will do!
Hubris? I'm officially in on this, and a new subscriber. I love hubris!
Welcome - hubris is a specialty.
In your research of existing cars roll centers to help decide your geometry, did you also see a pattern between roll center height and track widths? If your reference cars actually followed a 'rule'? You did not directly mention anything about that here. The closest is the camber curve which, of I remember correctly, helps to keep the COG over the roll center. I know it is impossible to cover everything in this length of video but curious as to if that was part of your process. Your classic car is narrower and taller than modern high performance cars so how much should your reference them? This probably could be a video in itself.
Interesting question. Track widths are wider these days and somewhat variable between car classes. The RCH has to be set in context to be sure and a value between 40 to 80 mm was pretty common. There are differences in how RCH is measured (force-based vs. geometric) and most of the forum material posted is from guys in their garage trying to sort things out (like me - and subject to uncertainty). What I found was they fell in that range. There is also variations in centre of gravity to consider. I'm unsure how closely I should follow - but there is variation in heights and there isn't a single value - same with camber curves. The Volvo had almost no camber gain or loss in it's original form - so guessing at a roll angle and then setting a camber curve is guess work. I had to pick one - I have raced cars with a lot of roll and it's not confidence inspiring - even on the road it's not what I prefer. I picked this rate based on this very limited experience.
loved it. i am still a little confused but thats not unusual, the ah-ha moment will come after a few read troughs.
This is well worth a subscription.
Thanks - more to come as we sort out some of the details I had to leave out of this first video.
What made you choose to start with a c4 front suspension? I don't think I would of started with a double a arm type.
I honestly thought it would be better than it is. I looked around and I could buy a cross member that I thought would fit and I was told by the company that this would be a nice solution. I can chalk this up to not wanting to do this. I thought I could buy something off the shelf and now I'm paying that price.
Awsome video. Thanks
Glad you liked it!
Great Video and explanation Sir♥♥
can you tell the name of the CAD software you used in this video?
Thanks. I used Fusion.
Great work well explained
This should help me
Put me down for a triple like
Thanks - glad it helped.
Can you list some of the books you've read on suspension design? I also want to read all the books.
nvm, I saw them flash by at 25:00...
I would add to this list Performance Vehicle Dynamics by Balkwill and Race Car Design by Seward - both excellent in their own right.
Are you sure your evaluation that test of the stock camber curve is correct ? At about 7:30 on. It looks to me like the wheel is going more neg camber with bump. Just eyeballing it compared to the background. Also your digital level has a generated image of a bubble level on it ? Its going to the right also, indicating the level is leaning to center of car. Or maybe im dyslexic .
I just went and checked it again and the stupid gauge was hung up on the lug nut. Dammit. I moved it a little and now the bubble is going the right way for the initial reading. The model in the computer is correct - the suspension in roll is losing contact patch. So what I said was right - what is shown is a bit of a mess.
@@ThrottleStopGarage Got ya ! So heres a question I have, regarding inputting suspension pivot pints into a computer model. In a situation like this, where the UCA pivot axis is inclined a fair bit downward (Anti dive), what point along that axis do you use for your 2D simulation ? The front ? The back ? its quite a difference in height above the ground plane.
Love your videos. Thanks for the work. I am a suspension tinkerer also.
@storminnorman1932 Think of it as a force vector, so take the middle elevation and use that point. It's the best you can do in 2D.
Fascinating
Thanks
Thank you for explaining where some of the points in space come from and what is acceptable from a design point of view. From the way your thinking is progressing I am assuming you are going to machine your own uprights. From my experience (formula car racing) and research the suspension design starts with the upright, especially from the home workshop environment, as it is expensive to manufacture. It is usually more efficient to start with a production upright that has a minimum of compromises. Have you got one in mind? I am currently have the some of the same issues on my current project, a 77 Monza, which have crappy front suspension geometry, crappy uprights that don’t allow big brakes, good hubs, etc.
That's basically what I'm doing by defining the scrub and SAI first. These are upright variables. I'll probably get the uprights machined. Next up, side view...then I can sort that out.
I was having difficulty understanding why you picked the ball joint locations, but after some thought i believe the lower inner was dictated by where the rack would be and the uppers chosen for the camber curve desired. As an aside your basic geometry is almost the same as Carrol Smith shows in Engineer to Win.
@@randycollett1746 That's pretty much the case. I tried to locate the lower so that I would have clearance to the brake rotor (this was a challenge) and had it at around the same height as the LCA mount on the frame. These took a few tries to get right. I've read all Carrol Smith's books - while I didn't follow his method in this video - it's more or less Ch 17 in RCVD, I'm not surprised that they're similar. Things for sure gelled for me and a lot of this stuff started to make sense very recently.
Great video Craig. I'm wondering if something like a modernized Mustang II/Pinto would have worked. Seems to be pretty popular with hot rodders. Not sure if the geometry is there or not but it looks decent at a glance. Looking forward to the next vid 👍
I am sure it can work. I can't find out a lot about it, and I'd rather just stop hacking and adapting parts while not understanding.
@@ThrottleStopGarage If you have the skill and resources like you do a clean sheet bespoke design is the best answer for sure. Plus the cool factor is way higher :)
@@jesscneal if I can get it to work...fingers crossed.
@@ThrottleStopGarage 🫰🫰
Are you still going to use the C4 vette spindle/upright or is that going to be made custom as well? I had planned on using the same setup in a 69 triumph GT6 but now I’m thinking it may not work since I need to narrow it so much. Great job on explaining your thoughts.
That upright is going to go and be replaced with something that I'm working on right now. Sorry for being so slow - this is pretty complicated stuff.
@@ThrottleStopGarage it seems very involved. I’m working on a triumph GT6 and trying to adapt a C4 suspension setup so am really enjoying the build. Thank you.
Another excellent video Craig. Are you trying to design a suspension that won't need adjusting, except for toe. Seems to me once you start adding shims or rotating cams all these parameters will change. Minutely, but will it be enough to blow your mind?
I think there has to be adjustability Alan - otherwise I'm going to get in trouble. I don't imagine that it's going to be right without the ability to move it around a bit.
@@ThrottleStopGarage Of course, I was just pulling your chain, didn't want your head to get too big, haha! Like everyone else here I'm in awe of your process. It's gotta be right or it's getting done again until it is. I saw a picture of you when I looked up the word patience in the dictionary, there you were!
No need to be looking at values from other cars except for sanity checks. Ask yourself what 'roll center height' actually means.
You're essentially designing an 'anti-roll' feature; and that's a percentage, not some best practice value in millimeter or freedom parsecs. You are determining how much leverage you want to give the mass attached to the CG.
Roll center on the tarmac means 0% anti-roll and a fun boat ride, RC at CG height means 100% anti-roll and zero roll even in a 5G turn.
So you want to figure out the necessary amount of geometrically build-in 'anti-roll' to get x degrees of roll at x Gs of max lateral acceleration.
The roll gradient for a boring family car is around 5°/G, a production sports car should be around 3-4°/G and little FSAE race cars are set up somewhere between 0.5 and 1.2°/G.
My suggestion would be a target of 1.3G and 3°/G, pretty sporty for a road car. So the starting point for simulations I'd pick would be 4° of max roll, 0.25-0.5° static camber and 4.25-4.5° of camber gain at max roll. It's not a GT3 car after all. For the rest you'll need weight, weight distribution and spring rates.
Get your spring rates sorted, calc the respective instant center heights, jacking forces, and see how much the car jacks up and rolls vs. what you want it to do, then apply the necessary 'anti-roll' by raising the static roll center to the right percentage of the CG height.
CPI btw has a lot of interplay with caster, it basically increases (neg.) camber on the outside wheel and reduces it on the inside wheel. The right angle again depends on the amount of jacking, roll and positive camber gain at full droop. The good thing is, you can leave this parameter for last since you can simply slide the upper control arm along its virtual construction line once everything else is locked in.
Hope this isn't too confusing, English isn't my first language.
Thanks for taking the time to write this out. I've gone over it several times - you've shared some very important thoughts. The RCH part is completely correct - my limitation is that I wasn't sure what that value should be for the percentage anti feature. The car originally had a RCH at ground level due to parallel (or close to it) control arm geometries and it was no fun. The method of "fixing" the suspension was to attach monster anti roll bars (25 mm - short arms) and increase the front spring rate. It was like a gocart. It lacked mechanical grip and I hated it.
What I've been missing to do the calculations is a sense of what a reasonable G force would be - even the other parameters are more me guessing and I hate that. This is what is holding up the side view work right now - how much anti-dive is reasonable? The caster is the same issue - there is a set value (sanity check if you like) that seems to pair with a SAI. So for a SAI of 12° - I see a caster of around 5° where for less SAI, I've observed more caster. I need to compute that rate.
Any advice on the side view construction parameters would be great. I understand that this is just "roll" in the pitch direction and that 100% would be alarming to drive as it would instantly react without any signaling to the driver. I think I've seen less than 20% tossed about as a start point but I have no idea if that's a good idea.
You’re describing camber loss while we watch the angle decrease and the bubble shows you’re getting gain. I’m confused.
Sorry...yes, it's gaining positive camber instead of gaining negative camber.
I notice that too! At first I thought I didn't know how to read my Klein Tools level (the bestest one ever!)
@@ThrottleStopGarage From my viewpoint you were gaining negative camber as the suspension was compressed....Adding spacers to the top arm imho will increase static positive camber.
This is just because of the camera perspective angle not being perfectly horizontal guys...
You can clearly see at ride height it had 0.4 degree negative camber and at 1 inch travel it shifted the other way to 0.5 degree positive camber, which is the wrong direction. It should have been 1.5 degree negative for example
This threw me off a bit too as visually it looked like it was gaining negative camber. The arm lengths and angles tell the story though. The bottom arm is past center of travel and will lose effective length while the upper are is rotating into a longer effective length.
Hi Craig: You can check out my approach anytime. Get a bunch of random suspension parts from China, throw them in by eyeball and hope for the best. Not a lot of calculation since we are measuring to thin air anyhow. The electric car driveline has arrived.
I'll have to drop over and see the electric parts!
I’ve been following you for a very…very long time, and I’ve got to say, this one so far is my favorite.
Thanks, much appreciated.
Have you watched "Fanatic Builds"? I think he ended up designing his suspension from scratch. Quite involved and interesting in any case. P.S. Love your vids.
Yes - I've seen some of his work. I will have to check out his approach.
Someone else: If it's good for Corvette its good enough for me.
You: Something's not kosher! Back to the drawing board. But first back to school.
Occupational hazard. I'm deeply uncomfortable with not understanding things.
It seems like it's going in a better direction, the corvette suspension looked wierd to me as it had the upper control arm sloping outwards rather than inwards as is custom, therefore the corvette instant centre would've been outside the wheel. I wonder what the result of mounting the inside of the corvette upper control arm lower would've been, this would to my mind put the instant centre in a more conventional spot (therefore a better roll centre) and also give better negative camber gain.
When you rolled the vehicle in the model it was still losing contact patch, perhaps it could use some more negative camber gain? Where is the steering rack in this? If you don't consider the location of the rack you'll likely get bumpsteer or rollsteer issues (you measured the width so I assume it was in the modeling).
Rather than inputing standard figures for camber gain why not calculate what camber you need? Camber gain required is equal to the degrees of roll, you've got the track width and a reasonable idea of suspension travel which should be enough to calculate it with some trigonometry. An older euro vehicle will tend to be less track width so may not be the same as more modern. Also camber curves depend on which suspension it is, macpherson struts go entirely the wrong way and are forced to start with too much negative camber so they have some left in bump.
Yes - the IC's are outside the wheel. Things are not quite as I've shown them as the level was not flat on the hub and eagle-eyed viewers picked up what I was too tired to notice. The computer model is correct, it's a little not what you'd expect - point made. When I optimized it before, the 3D software had me moving a lot of things to get the attributes I wanted and it looked quite different. That's why I bought the stock crossmember - to find out what GM thought it should be. Steering rack is in place in the 2D model and designed around zero bump steer - probably the most important parameter.
I'm just learning - the trick for me was not knowing what "reasonable" is with the parameter. So what's a reasonable lateral G? What sort of roll is not crazy? How little is too little? I'm all ears for feedback on these parameters. They are what is needed to do the entire thing properly. I am going to do this for the 3D model - but the 3D model needs inputs and I just needed to start. Does this make sense?
@@ThrottleStopGarage a lot of even amateur racers have lateral G's on thier videos, most seem to see less than 1.5G from memory. When I calculated roll for my suspension I used available travel in bump and droop from ride height, ie if you've 4" inches in bump and 4" inches in droop that's 8" across the track width of the car trigonometry means you can work out an angle of roll that would be the upper limit.
1:40
❤
I find myself in too many diy projects because nobody’s invented it yet
6:22 LOL
Horses for courses.
Indeed - I find this teetering on the edge of the unknown stuff oddly relaxing. I think we're just wired to innovate.
Although automotive companies spend millions on their designs, they have a lot more constraints than you do. You're building one custom vehicle for your own consumption, not trying to sell millions of vehicles to the masses.
I realize you don't want to use an existing design, however you have the ability to leverage the knowledge others spent creating existing designs. This allow you to choose the benefits and compromises you're willing to accept based on your own criteria, parts availability, and packaging.
I hope you see success on the suspension design and you're able to build something to surpass your expectations.
Thanks, the positive energy is much appreciated! I hope it meets my expectations - if not, we'll do it again!
@ThrottleStopGarage That is the attitude to have!
which software did you use?
I used Fusion
❤️🔥