Loving the progress. Every time I see the image of the body it shocks me how similar our designs are. I got sucked into boats but hope to pick back up on the car project this winter.
Hey this seems like a very nice project indeed! Obviously it seems I may be somewhat late to the party. Since the front and rear suspension allow for no roll movement of the tire relative to the vehicle, at the moment there exists no roll axis on your vehicle. There is (approximately) no body roll and as soon as the maximum tilt angle is surpassed the vehicle suddenly changes behavior and thus becomes unstable. As an idea I would try to make transition between stable and unstable behaviour as smooth as possible. Generating roll compliance of some form in order to have a time delay for the rollover event itself seems like a possible successfull strategy to me. The idea being to give your controller actually some time to give correcting inputs before the vehicle tips over. For the rear axle with e.g. a (Motorcycle-style) rounded sidewall rear wheel you may loose some maximum tilt angle, but prevent the sidewall of the tire "hooking up" when rolling over or skipping sideways over the surface and generally making the rear wheel less sensible to body roll. On the front maybe increasing the flex with a single pivot suspension similar to a torsion beam axle might be an option to get some compliance in the roll axis. Since your main target is to drive fast in a straight line, maximum tilt angle might actually not be the main target. If you think about it even a Motorcycle (or Bicycle) is capable of self-stable driving. So I guess the most radical solution could be to reduce roll stiffness to a minimum and make a motorcyle out of it (then you would probably need to push it at the start). In that case I would suggest increasing the mechanical trail (always a good idea for straight line stability) and give the steering axis some inclination. In the end you don't need it to be able to turn sharp corners, you just need it to not turn sharply.
Wow, thank you for all the feedback. Especially about the roll axis. I never was sure of the exact dynamics of roll since the car didn't really match anything else. After that last video, I slowly realized that this car is just a 3 wheel motorcycle. I'm working on version 3 but I'm still not sure how much caster to include. I tried 15 degrees caster initially, but the wheel's gain too much camber when turning and I can't get enough steering response, so I'm going to settle for around 5-10 degrees and see how it works. You're completely on the mark with the sidewall of the tire hooking up. I can actually see chunks torn out of the tire when it spins out or rolls over. I've since switched to a much softer tire rear tire, almost a wrinkle wall slick. This may or may not hold up at high speeds, but makes the car much more stable at low speeds. I'm actually questioning if I need a suspension at all with a soft enough tire.
I think the rough surfaces you're running on compared to the diameter tire would be like running a full size vehicle over speed bumps. Have you tried this in a school gym or some other incredibly smooth paved surface? Can you increase the wheel diameter and raise the axel height, leaving you at the same ride height, but with a higher roll point and a smoother ride?
I'm so glad to see the series continue! If it were my project, I'd probably find a way of widening the track width. From your data, it looks like a large improvement of the roll/tip angle can be had for a small increase in width. I remember one of the prime goals of this project was to keep the correctional area as low as possible to minimize drag, but if it isn't stable, you're not gonna go fast anyway. So maybe widen the track width until you get a roll/tip angle in the 40 to 50 degree range? You helpfully wouldn't need more stability than that and it also would hopefully not widen the car too much.
I agree, it shouldn’t take much extra width to stabilize the car, but it makes me sad because frontal area grows so quickly. Realistically, at this size, I could even get away with just a set of faired wheels separate from the main body.
If you put the wheels far enough apart and suspend them independently, you could move the battery forward to sit somewhere between the front wheels. That would improve stability as well since reverse trikes have inherent instability (ask any tail dragger pilot). The obvious downside there would be added mass but it’s something to think about.
When front wheel is in the air after some bumps on high speed, and rear wheel accelerate whole machine, it is uncontrollable in this moment. So need wing in front. Also you need some spoiler or use groundeffect to get more stability. You should control roll with that wings
I agree. Not having a front wing may cause me some issues and I may move away from wings completely and run a ground effect floor. It's hard to see on camera, but I run a significant amount of rake, meaning the rear is always higher than the front to maximize the front downforce from the floor. CFD shows that there is more downforce on the front than the rear with the wing, but sadly I don't have a windtunnel to confirm that this is the case.
Awesome work! I've done a lot of random RC builds based on aero and I'm finally getting my design together for a speed running car. My goal is to have really efficient aero....ive considered doing active aero to weight balance front rear but may settle on CFD work. I'm not sure what your speed goal is, but you're limitation looks to be traction with the TPU wheel and limited rear downforce. Have you considered a stickier tire?
Thanks! I loved your fan car and active aero videos. Speed running cars are a lot of fun to design. I really love the idea of active aero and I will revisit it later. Mine is long enough, front weight biased, and high raked that it’s relatively aero stable. The rear tire structure is TPU but it’s capped in soft rubber from a mountain bike inner tube. Kind of ghetto, but I had a hard time finding an appropriately sized RC car tire. I’m building a tire test rig right now to test EVA foam and other materials. I think the limiting factor is finding a place to run. Having more traction for acceleration and stopping will help significantly. I saw a full size drag car that did 212mph in the 1/8mi and that’s my new goal 😆
@@IndeterminateDesign have you done the math to see what your tractive force is compared to wind resistance at your target speed? I would imagine you're lucky to get a practical friction coefficient of 1, which means less than 50% of your total mass is your baseline traction force depending on weight balance. So your practical acceleration at low speeds is limited to like 0.5 Gs. Then when your have reasonable downforce, you have the drag + inertial resistance to acceleration to contend with. But that means you get more acceleration at higher speeds if your motor has enough torque to take advantage of it at those. This is one of the reasons it's so common to use AWD instead of rear drive (all weight and downforce goes to traction). Funny you mention dragsters. My build is based on a dragster chassis. Getting over 200mph on a 1/8th take over 2Gs of average acceleration which I would consider high for an ultra low drag car.
I don’t have a lot of confidence in my CFD simulations, but the aero drag is much lower than I expected, typically 6-9N at 200mph with .005m² of frontal area. In theory, that could be overcome with little to no downforce on the rear. I’m sure you’ve read all of the magazine articles about Nic Case’s 202mph setup. I did some back of the napkin calcs based on his frontal area, weight, and power and my setup should exceed what he had. That's really the only comparison data I've been able to find. I wish I could package and AWD setup. My first car took over ⅓ of a mile to hit 90mph because it could so easily lose traction or lift the nose and take flight. Given unlimited distance and radio range I think hitting 200mph would be pretty straight forward. The max acceleration I recorded with my first car was 1.2G acceleration from 0-50mph. When I saw your video with the fan car, I kept thinking my numbers just don’t seem right. Studying drag cars, I’m seeing some crazy acceleration figures even on drag rails without rear wings, but I don’t know if a drag slick would hold up to 25,000+ rpm. An expanding wrinkle wall slick would be perfect for our application if it could take the RPM. It’s great getting to talk technical. If you ever want to reach out or have any questions my email is indeterminatedesign@gmail.com
@@IndeterminateDesign whoa! .005m^2?! That's way smaller than I thought it was. That's like 7x7cm square! My current plan is .035m^2, but probably way more power as my battery has the same frontal area as your car. Dragsters get their insane acceleration with track prep and massive contact patches at launch. 250in^2 glued to the ground can support a lot of tractive force. And then something like a top fuel dragster has thousands of pounds of downforce before the mid point of the track. Dragsters also rely heavily on weight transfer to the rear axle. I'm not sure I'd trust the 1.2 G longitudinal acceleration at low speed as physics dont really support that without downforce. Maybe you're getting some readings from surface roughness. Best of luck and I look forward to the design of this thing coming along. If all else fails...let's slap rockets on these things and go to a drag strip. Thanks for the email...once I finish these next 2 projects I'll be focusing on the speed car and will reach out!
You can stabilize the car by using two motors with differential throttle with a gyro and making hub motors out of low kv motors. The front steering is not good enough you need that differential throttle for the car to tack straight because the front wheels will just skip across the ground without steering the car.
That makes sense. The car seems to steer less and less with speed. I am curious to at least play with a gyro setup to see how much force is actually needed to keep it straight.
I’ve thought about this. I actually read some papers on electric cars where they used the battery pack as a tuned mass damper. I’d need something like soft foam to wrap the battery in.
@@IndeterminateDesign Dont need soft foam for that. Suspend the batter with rubber bands, then dampen it with pieces of foam at the top and bottom to tune the mass, and prevent oscillating. But first, fix the suspension geometry. Thats most of your issues.
There are a bunch of things to do differently here. First, without getting WAY to in depth on the geometry of roll centers, your "axle centerline" is not the roll center. Its actually largely dependent on your suspension mounting position. For example, adding a 1mm spacer under the ball stud of my 2WD off-road buggy, is enough roll center adjustment to take the car from traction rolling when looking at every corner, to providing enough camber gain to allow the car to get just enough grip at speed to comfortably make the corner. In short, your rear suspension arm is too damn high. It needs to be mounted at the absolute bottom of the chassis And angle UP to the rear tire. This will SIGNIFICANTLY drop your roll center, and Center of Gravity. Additionally, your front end geometry is severely lacking in trail, or caster. I suggest you look into caster, rake, and trail, and get a basic understanding of how they work on a car and a motorcycle, and how that accommodates self centering steering geometry. That way the car wants to keep straight, and doesnt continually produce a pitching moment that throws the car sideways. And...just like the rear, your bottom most pivot point of yoru front suspension should be in the absolute bottom of the chassis. Not the middle like it is now. These changes should also allow you to lower the whole body of the streamliner a couple mm. Being close to the ground should provide some stability benefit. Also you may consider the chance to play with some ground effect on the bottom as well. As this will greatly help the stability, without adding undue drag.
I’m not an expert, but I assumed the roll center would behave statically when there’s no role freedom. But it makes sense what you’re saying, and that probably means a softer suspension would help as well. I will try out a new rear mount here if I can. I am curious on the caster and trail on the front. I know they help in a real car where the wheels will self center on their own, but with a steering servo that maintains a set position does this still matter?
@@IndeterminateDesign the roll center exists whether there is roll freedom or not. Your tilt tests prove this. Roll center is a byproduct of suspension geometry, and how it interacts with the center of mass. Look at the difference in swing arm placement of a road racing bike, and a drag bike. One is to keep the roll center as centralized to the center of mass as possible to help the bike rotate axially to lean. While the other is underslung and extended to aid in traction, and straight line stability. Ducati even put a "hole shot" device on their MotoGP bike that is essentially a pneumatic shock that drops the swing arm for acceleration at the start, and raises back up for turning corners. Besides that, as soon as the vehicle starts to steer, the roll center will change the instant the moment of inertia changes. And all sorts of complex physics start to take place, and plop...over the car goes. As for the caster, yeah, it will help. Your steering servo can't activatly correct the slop on the steering, as quickly as A good caster geometry and the forces it generates can self correct it for you. The two together should make it even better. The more the car can do on its own to keep it tracking straight without any control input, the faster it will be. Unless you are intentionally trying to make it an "acrobatic streamliner", you should try to bake in as much inherent stability as possible, instead of trying to program your way around it with crutches and auxiliary systems.
Too much bumps on road, your accelerometer values and decision you made on this data will go mad. I think you should use stability wings and control whoke system with them instead of just front wheels.
I agree. I have to filter the inputs from the MPU to have a clean enough signal for the steering servo. I initially tried to lower the throttle based on the yaw rate, but the ESC I was using was from an airplane which caused the rear wheel to lock up. I have upgraded to an ESC meant for an RC car for Version 3 which works much better.
Welcome back, mabye adding some kind of gyros to increase stability might be helpful
Loving the progress. Every time I see the image of the body it shocks me how similar our designs are. I got sucked into boats but hope to pick back up on the car project this winter.
Awesome, can’t wait to see yours. It’s a tougher problem to crack than I thought it would be to design a car from scratch.
@@IndeterminateDesign yes absolutely. The difficulty and challenge is what makes it so much fun from an engineering perspective.
having a teardrop cowl to accomodate a much wider front wheelspan might help
Hey this seems like a very nice project indeed!
Obviously it seems I may be somewhat late to the party.
Since the front and rear suspension allow for no roll movement of the tire relative to the vehicle, at the moment there exists no roll axis on your vehicle. There is (approximately) no body roll and as soon as the maximum tilt angle is surpassed the vehicle suddenly changes behavior and thus becomes unstable.
As an idea I would try to make transition between stable and unstable behaviour as smooth as possible. Generating roll compliance of some form in order to have a time delay for the rollover event itself seems like a possible successfull strategy to me. The idea being to give your controller actually some time to give correcting inputs before the vehicle tips over.
For the rear axle with e.g. a (Motorcycle-style) rounded sidewall rear wheel you may loose some maximum tilt angle, but prevent the sidewall of the tire "hooking up" when rolling over or skipping sideways over the surface and generally making the rear wheel less sensible to body roll.
On the front maybe increasing the flex with a single pivot suspension similar to a torsion beam axle might be an option to get some compliance in the roll axis.
Since your main target is to drive fast in a straight line, maximum tilt angle might actually not be the main target. If you think about it even a Motorcycle (or Bicycle) is capable of self-stable driving.
So I guess the most radical solution could be to reduce roll stiffness to a minimum and make a motorcyle out of it (then you would probably need to push it at the start). In that case I would suggest increasing the mechanical trail (always a good idea for straight line stability) and give the steering axis some inclination.
In the end you don't need it to be able to turn sharp corners, you just need it to not turn sharply.
Wow, thank you for all the feedback. Especially about the roll axis. I never was sure of the exact dynamics of roll since the car didn't really match anything else. After that last video, I slowly realized that this car is just a 3 wheel motorcycle. I'm working on version 3 but I'm still not sure how much caster to include. I tried 15 degrees caster initially, but the wheel's gain too much camber when turning and I can't get enough steering response, so I'm going to settle for around 5-10 degrees and see how it works.
You're completely on the mark with the sidewall of the tire hooking up. I can actually see chunks torn out of the tire when it spins out or rolls over. I've since switched to a much softer tire rear tire, almost a wrinkle wall slick. This may or may not hold up at high speeds, but makes the car much more stable at low speeds. I'm actually questioning if I need a suspension at all with a soft enough tire.
I think the rough surfaces you're running on compared to the diameter tire would be like running a full size vehicle over speed bumps.
Have you tried this in a school gym or some other incredibly smooth paved surface?
Can you increase the wheel diameter and raise the axel height, leaving you at the same ride height, but with a higher roll point and a smoother ride?
Amazing project. It seems to be to suffer from the same thrust problem as missiles. How do you think to add some winglet?
I'm so glad to see the series continue! If it were my project, I'd probably find a way of widening the track width. From your data, it looks like a large improvement of the roll/tip angle can be had for a small increase in width. I remember one of the prime goals of this project was to keep the correctional area as low as possible to minimize drag, but if it isn't stable, you're not gonna go fast anyway. So maybe widen the track width until you get a roll/tip angle in the 40 to 50 degree range? You helpfully wouldn't need more stability than that and it also would hopefully not widen the car too much.
I agree, it shouldn’t take much extra width to stabilize the car, but it makes me sad because frontal area grows so quickly. Realistically, at this size, I could even get away with just a set of faired wheels separate from the main body.
If you put the wheels far enough apart and suspend them independently, you could move the battery forward to sit somewhere between the front wheels. That would improve stability as well since reverse trikes have inherent instability (ask any tail dragger pilot). The obvious downside there would be added mass but it’s something to think about.
Thank you i would make it a bit wides in the fron also the rear whell i would make it wider too just my ideas and or my 2 cents
When front wheel is in the air after some bumps on high speed, and rear wheel accelerate whole machine, it is uncontrollable in this moment. So need wing in front.
Also you need some spoiler or use groundeffect to get more stability. You should control roll with that wings
I agree. Not having a front wing may cause me some issues and I may move away from wings completely and run a ground effect floor. It's hard to see on camera, but I run a significant amount of rake, meaning the rear is always higher than the front to maximize the front downforce from the floor. CFD shows that there is more downforce on the front than the rear with the wing, but sadly I don't have a windtunnel to confirm that this is the case.
Awesome work! I've done a lot of random RC builds based on aero and I'm finally getting my design together for a speed running car. My goal is to have really efficient aero....ive considered doing active aero to weight balance front rear but may settle on CFD work.
I'm not sure what your speed goal is, but you're limitation looks to be traction with the TPU wheel and limited rear downforce. Have you considered a stickier tire?
Thanks! I loved your fan car and active aero videos. Speed running cars are a lot of fun to design. I really love the idea of active aero and I will revisit it later. Mine is long enough, front weight biased, and high raked that it’s relatively aero stable.
The rear tire structure is TPU but it’s capped in soft rubber from a mountain bike inner tube. Kind of ghetto, but I had a hard time finding an appropriately sized RC car tire.
I’m building a tire test rig right now to test EVA foam and other materials. I think the limiting factor is finding a place to run. Having more traction for acceleration and stopping will help significantly. I saw a full size drag car that did 212mph in the 1/8mi and that’s my new goal 😆
@@IndeterminateDesign have you done the math to see what your tractive force is compared to wind resistance at your target speed?
I would imagine you're lucky to get a practical friction coefficient of 1, which means less than 50% of your total mass is your baseline traction force depending on weight balance. So your practical acceleration at low speeds is limited to like 0.5 Gs. Then when your have reasonable downforce, you have the drag + inertial resistance to acceleration to contend with. But that means you get more acceleration at higher speeds if your motor has enough torque to take advantage of it at those. This is one of the reasons it's so common to use AWD instead of rear drive (all weight and downforce goes to traction).
Funny you mention dragsters. My build is based on a dragster chassis. Getting over 200mph on a 1/8th take over 2Gs of average acceleration which I would consider high for an ultra low drag car.
I don’t have a lot of confidence in my CFD simulations, but the aero drag is much lower than I expected, typically 6-9N at 200mph with .005m² of frontal area. In theory, that could be overcome with little to no downforce on the rear.
I’m sure you’ve read all of the magazine articles about Nic Case’s 202mph setup. I did some back of the napkin calcs based on his frontal area, weight, and power and my setup should exceed what he had. That's really the only comparison data I've been able to find.
I wish I could package and AWD setup. My first car took over ⅓ of a mile to hit 90mph because it could so easily lose traction or lift the nose and take flight. Given unlimited distance and radio range I think hitting 200mph would be pretty straight forward.
The max acceleration I recorded with my first car was 1.2G acceleration from 0-50mph.
When I saw your video with the fan car, I kept thinking my numbers just don’t seem right.
Studying drag cars, I’m seeing some crazy acceleration figures even on drag rails without rear wings, but I don’t know if a drag slick would hold up to 25,000+ rpm. An expanding wrinkle wall slick would be perfect for our application if it could take the RPM.
It’s great getting to talk technical. If you ever want to reach out or have any questions my email is indeterminatedesign@gmail.com
@@IndeterminateDesign whoa! .005m^2?! That's way smaller than I thought it was. That's like 7x7cm square! My current plan is .035m^2, but probably way more power as my battery has the same frontal area as your car.
Dragsters get their insane acceleration with track prep and massive contact patches at launch. 250in^2 glued to the ground can support a lot of tractive force. And then something like a top fuel dragster has thousands of pounds of downforce before the mid point of the track. Dragsters also rely heavily on weight transfer to the rear axle.
I'm not sure I'd trust the 1.2 G longitudinal acceleration at low speed as physics dont really support that without downforce. Maybe you're getting some readings from surface roughness.
Best of luck and I look forward to the design of this thing coming along. If all else fails...let's slap rockets on these things and go to a drag strip.
Thanks for the email...once I finish these next 2 projects I'll be focusing on the speed car and will reach out!
You can stabilize the car by using two motors with differential throttle with a gyro and making hub motors out of low kv motors. The front steering is not good enough you need that differential throttle for the car to tack straight because the front wheels will just skip across the ground without steering the car.
That makes sense. The car seems to steer less and less with speed. I am curious to at least play with a gyro setup to see how much force is actually needed to keep it straight.
Use much bigger wheels to move the axis of rotation higher from the center of mass.
Look into renault F1 tuned mass damper from their f1 cars with alonso, it made a huge difference fro them
I’ve thought about this. I actually read some papers on electric cars where they used the battery pack as a tuned mass damper. I’d need something like soft foam to wrap the battery in.
@@IndeterminateDesign seems like it might work, love the preoject
@@IndeterminateDesign Dont need soft foam for that. Suspend the batter with rubber bands, then dampen it with pieces of foam at the top and bottom to tune the mass, and prevent oscillating.
But first, fix the suspension geometry. Thats most of your issues.
There are a bunch of things to do differently here. First, without getting WAY to in depth on the geometry of roll centers, your "axle centerline" is not the roll center. Its actually largely dependent on your suspension mounting position. For example, adding a 1mm spacer under the ball stud of my 2WD off-road buggy, is enough roll center adjustment to take the car from traction rolling when looking at every corner, to providing enough camber gain to allow the car to get just enough grip at speed to comfortably make the corner.
In short, your rear suspension arm is too damn high. It needs to be mounted at the absolute bottom of the chassis And angle UP to the rear tire. This will SIGNIFICANTLY drop your roll center, and Center of Gravity. Additionally, your front end geometry is severely lacking in trail, or caster. I suggest you look into caster, rake, and trail, and get a basic understanding of how they work on a car and a motorcycle, and how that accommodates self centering steering geometry. That way the car wants to keep straight, and doesnt continually produce a pitching moment that throws the car sideways. And...just like the rear, your bottom most pivot point of yoru front suspension should be in the absolute bottom of the chassis. Not the middle like it is now.
These changes should also allow you to lower the whole body of the streamliner a couple mm. Being close to the ground should provide some stability benefit. Also you may consider the chance to play with some ground effect on the bottom as well. As this will greatly help the stability, without adding undue drag.
I’m not an expert, but I assumed the roll center would behave statically when there’s no role freedom. But it makes sense what you’re saying, and that probably means a softer suspension would help as well. I will try out a new rear mount here if I can.
I am curious on the caster and trail on the front. I know they help in a real car where the wheels will self center on their own, but with a steering servo that maintains a set position does this still matter?
@@IndeterminateDesign the roll center exists whether there is roll freedom or not. Your tilt tests prove this. Roll center is a byproduct of suspension geometry, and how it interacts with the center of mass. Look at the difference in swing arm placement of a road racing bike, and a drag bike. One is to keep the roll center as centralized to the center of mass as possible to help the bike rotate axially to lean. While the other is underslung and extended to aid in traction, and straight line stability. Ducati even put a "hole shot" device on their MotoGP bike that is essentially a pneumatic shock that drops the swing arm for acceleration at the start, and raises back up for turning corners. Besides that, as soon as the vehicle starts to steer, the roll center will change the instant the moment of inertia changes. And all sorts of complex physics start to take place, and plop...over the car goes.
As for the caster, yeah, it will help. Your steering servo can't activatly correct the slop on the steering, as quickly as A good caster geometry and the forces it generates can self correct it for you. The two together should make it even better.
The more the car can do on its own to keep it tracking straight without any control input, the faster it will be. Unless you are intentionally trying to make it an "acrobatic streamliner", you should try to bake in as much inherent stability as possible, instead of trying to program your way around it with crutches and auxiliary systems.
Too much bumps on road, your accelerometer values and decision you made on this data will go mad.
I think you should use stability wings and control whoke system with them instead of just front wheels.
I agree. I have to filter the inputs from the MPU to have a clean enough signal for the steering servo. I initially tried to lower the throttle based on the yaw rate, but the ESC I was using was from an airplane which caused the rear wheel to lock up. I have upgraded to an ESC meant for an RC car for Version 3 which works much better.
"nice smooth asphalt" Ok, so you're nowhere near Phoenix then.
It’s a rarity to find anywhere anymore.