Awesome progress! As a mechanical engineering student I appreciate the technical explanations which unfortunately many other youtubers exclude from their videos. Keep it up 👍
Great progress and love the design. Good roads make all the difference. Florida has the best I have seen. I also found a good one in Oklahoma that I drive to now. It's a 2hr round trip but worth it.
Impressed by the evolution, really like how you developed this car. I think that a big part of down force is generated by the rear sidepods of the car that act as "wing" and not so much by the ground effect underfloor. The surface area that can generate downforce underneath the car is very small (small portion between throat and kick point), that 's why I am saying that the ground effect is minimal. The road roughness is not helping either the airflow underneath the car. I am trying to build some sort of streamliner myself, but the issue that I have is that at high speeds the suspension is loaded and even small bumps in the roads disturb the car in a such way that I have to abort the run. I tried to reduce the downforce to unload the suspension but still I don't have consistency. After the winter I will try to make some active aero to make the car float to keep the pitch between limits and have "just enough" downforces. What is the width of the car? I asked you on an old video, did you tried to run it as a FWD car? Maybe will improve stability during acceleration and braking.
Thanks. Wouldn't the diffuser tunnel should continue to generate downforce due to the increased speed of the air past the throat? The CFD results I saw from other cars showed this. I agree, the texture ground surface can greatly lower downforce. I tried this on fresh asphalt vs concrete and the difference was noticeable. Also just changing the boundary conditions in OpenFOAM halved the downforce when the surface roughness was increased.
@@IndeterminateDesign The diffuser will generate minimal downforce, after kick point you increase the height but also in width of the diffuser as the body is chamfered inwards and this cause the air velocity to slowdown to match the speed of the air from rear of the car. A think that the slope is way to high and even with the turbulent airflow (to delay the separation point), the separation may occur earlier than is supposed to. Even though the body has great influence, a safe rule of thumb is for the diffuser to have between 10 and 15 degrees slope.
I forgot to answer on the first one, the streamliner is 80mm at the front and the widest part of the diffuser is 120mm. I think I would need a small wind tunnel or something to see what's going on with the diffuser. It's definitely working, but I don't know by what mechanism. The diffuser kick up is 9 degrees, and I went with a contoured throat to help with height sensitivity. It doesn't look it, but most of the expansion is in the horizontal directly as well. I've added vortex tunnels to the edges of the latest diffuser design which the CFD shows increased down force. Hopefully if I can get more high speed runs I can see the data logging trends and measure the increase in traction. I did try running the car in reverse. I flipping the front suspension upside down to have the right caster. I found it was stable at low speeds
@@IndeterminateDesign Thank you for the answer. Something about the simulation doesn't look right, the lowest pressure is way over the kick point (where you expect to see it), almost to the middle of the diffuser where the air is slowed down. Usually you expect that on the throat zone to have the high air velocity but from what I am seeing at 5:18 this is not the case. You can search "Aero analysis for F1 2022 based on CFD result" and see the lowest pressure area of an F1 car.
Amazing work with each video !! Which software you used for aerodynamics simulation ? And is the car all in pla ? Would you fear the scenario where you forgot it in the car while the car is in the hot sun which deform it due to high temp inside your car ?
I’m using Simscale for a lot of the “rough” CFD work and I’m using BlueCFD (OpenFOAM for windows) for the fiber meshing. The body of the car is all PLA, the wheels and suspension are all carbon nylon or carbon polycarbonate. I use PLA for my prototypes, and I was in a rush and decided to try to run the PLA wheel and it wasn’t great. I’m pretty careful to no leave the streamliner in a hot car. It’s actually PLA+ and it has a bit higher melt temp than normal PLA. I have had the ESC make the top of the monocoque soft if I do back to back runs. I may add vents for testing. Eventually the whole thing will be printed in polycarbonate to save weight, but I haven’t gotten to that point yet.
Hopefully constructive opinion: 1. Forget all the aero stuff. You need a controllable reliable car first. Focus on that. 1a. Get a good GPS module to log direction/speed/acceleration/deceleration. (Crashing) 2. Most high-speed R/C cars use a type of foam drive tire. Use that first to get a driveable car, then improve on what they have done. 3. You should be able to calculate the rear tire speed at 100mph. Run that on the bench to measure tire growth and reliability. Save that RPM for later. Log.. 5. Examine how fast the rear tires are actually spinning when the GPS says you are at 100mph. Compare to estimates to get tire slippage. A big issue with other R/C cars. Log 5. Measure your ESC power usage at 40,50,60,100~ to create a power/speed profile. This will help you develop your aero package. Log. 6. Take a look at modern Drag Racing bikes to get an idea of what the "profile" of the rear tire should be. They have been doing it for years. This will help driveability. Should be the top focus. 7. Put the car down, and drive 5 feet in the direction you want the car to go. Program the steering to follow data from the GPS to follow that direction. Let the car correct itself instead of you trying to drive. This will need lots of learning/programming as you will want the steering to be less sensitive as the car speed increases. Fun!! 8. Use the GPS or other monitoring device to slowly remove the throttle when slowing down. This should help your braking issue. 9. Use a piece of masking tape on the edges and along the bottom of the car. It will show where the car is hitting the ground and it is easily replaced. 10. Determine ride height at different speeds for the front and rear. Use temporary stick-on winglets to adjust the ride height or downforce in the field for estimated body shape changes. 11. The skis should be taped onto the car's bottom. This allows field adjustment to correct for too high/log ride height in the field. Use shims or tape to change the angle of attack! 12. Outboard skis are da bomb. Possibly replaceable PLA nubs/bumpers. Engineer in a quick swap nose as it's first to "wear out". Bring a couple. That's what F1 guys do. 13. Build the swing arm so you can adjust the driving direction in the field. With more power, you don't know where the car is going. I could go on, but you need to have some fun also. I think the idea is to build a car that is fast/drivable first, measure/log everything, then start making go-fast parts. Having lots of shims/tape/spacers etc. in the field is super important to make adjustments while testing. And maybe get a cheap tripod. thx, John
Thank you for all the really well thought out ideas and feedback! 1. I agree. I think this is the most controllable version of the car I've had so far. I really am tempted to wide the wheel base or try a 4 wheel design to make the car even more stable and easier to control. -I haven't tried to use the headings from the GPS yet. I think I would most likely need to combine the GPS data with the gyro/accelerometer data with a Kalman filter to have something responsive enough. I actually just found a fatal software flaw where the reading of the GPS data was randomly delaying the microcontroller and stopping all of the stability control systems. I've since switched to using the additional processor on the ESP32 for the GPS and logging, and this has dramatically improved the heading control. I really need multiple dedicated processors so I can separate dependencies. 2. I really want to switch to the foam type tire, but I haven't been able to find the right EPDM foam since I have to custom make my own tires since I can't find the unglued foam tires. It requires a higher durometer foam, and the foam I have tried chunked out at less than 80mph. 3. I admittedly haven't done enough tire testing. The rubber tires grow dramatically even with a homemade belt made of dyneema. The tires self clearance the car's body right now. That's a great idea to create a test rig to run the tire for a prolonged period of time. Do you think I need a roller to exert the force from the road? 5. I just added a rear wheel speed sensor to the car. I've also implemented traction control and antilock braking. Additionally I setup linear interpolation on the throttle and brake, which eases on the acceleration or braking. 5a. Characterizing the motor is something I've been working hard on. The traction control I'm using is rate based, which isn't ideal, but it is used by a lot of drag racers. Ideally, I'd have a motor dyno that allows me to fully model the throttle, RPM, and torque output of the motor. The RC car ESC's are not linear in their torque response. Unfortunately I haven't found an off the shelf solution for logging the motor current that's Arduino compatible. I may be forced to build my own. I am building a voltage logging circuit. I'm using a very small battery which is most likely sagging at higher speeds as well. 6. I have found a semi rounded rear tire helps stability. I think with a wider front width I could run a more square tire and have better traction. 7. I do the heading testing with the gyro and accelerometer and they do pretty well up to a point. I need to do more testing now that I've fixed the software bug. I think after 80mph the car is pretty much being driven by aerodynamics as I have virtually no control over it. 8. I've added the deceleration taper based on wheel speed but haven't tested it. I think a parachute will help to stabilize the car as well, I see this a lot on drag racers when they lose the rear end they pull the chute. 9. I like the masking tape idea, plus it adds a little more wear resistance. 10. I really want to log the ride height. Space is the main issue I have right now. I have some lidar sensors that might be useful if I can get the car more reliable. 11. That's a great point. I need much more adjustability. I was hoping to keep the design simple and just reprint parts, but like you said I need to make adjustments in the field. 12. I am trying some polyurethane skiis for the diffuser edges. It is a surprisingly durable material. 13. I agree on this. I have essentially a "spring rubber" setup for the rear swing arm. I just need to print out several variations and test them out. Like you said, making the car drivable is the most difficult part. It's hard to find the ideal conditions and perfectly smooth surfaces in which to run the car right now. Again, thanks for the feedback on this. I really enjoy trying to think through all of these possibilities.
i love designing, cant wait to see how my car compares to others on the rc speedrunning field
Awesome progress! As a mechanical engineering student I appreciate the technical explanations which unfortunately many other youtubers exclude from their videos. Keep it up 👍
Awesome! This thing is really getting the mature feel to it
133 is very impressive! I'm happy to see this project progressing well. The diffuser is a very cool addition.
Great progress and love the design. Good roads make all the difference. Florida has the best I have seen. I also found a good one in Oklahoma that I drive to now. It's a 2hr round trip but worth it.
Yeah, picking the right roads is key. That last run actually was in OK near Atoka. It ends in a hill which helped braking 😅
Impressed by the evolution, really like how you developed this car. I think that a big part of down force is generated by the rear sidepods of the car that act as "wing" and not so much by the ground effect underfloor. The surface area that can generate downforce underneath the car is very small (small portion between throat and kick point), that 's why I am saying that the ground effect is minimal. The road roughness is not helping either the airflow underneath the car. I am trying to build some sort of streamliner myself, but the issue that I have is that at high speeds the suspension is loaded and even small bumps in the roads disturb the car in a such way that I have to abort the run. I tried to reduce the downforce to unload the suspension but still I don't have consistency. After the winter I will try to make some active aero to make the car float to keep the pitch between limits and have "just enough" downforces.
What is the width of the car? I asked you on an old video, did you tried to run it as a FWD car? Maybe will improve stability during acceleration and braking.
Thanks. Wouldn't the diffuser tunnel should continue to generate downforce due to the increased speed of the air past the throat? The CFD results I saw from other cars showed this. I agree, the texture ground surface can greatly lower downforce. I tried this on fresh asphalt vs concrete and the difference was noticeable. Also just changing the boundary conditions in OpenFOAM halved the downforce when the surface roughness was increased.
@@IndeterminateDesign The diffuser will generate minimal downforce, after kick point you increase the height but also in width of the diffuser as the body is chamfered inwards and this cause the air velocity to slowdown to match the speed of the air from rear of the car. A think that the slope is way to high and even with the turbulent airflow (to delay the separation point), the separation may occur earlier than is supposed to. Even though the body has great influence, a safe rule of thumb is for the diffuser to have between 10 and 15 degrees slope.
I forgot to answer on the first one, the streamliner is 80mm at the front and the widest part of the diffuser is 120mm. I think I would need a small wind tunnel or something to see what's going on with the diffuser.
It's definitely working, but I don't know by what mechanism. The diffuser kick up is 9 degrees, and I went with a contoured throat to help with height sensitivity. It doesn't look it, but most of the expansion is in the horizontal directly as well. I've added vortex tunnels to the edges of the latest diffuser design which the CFD shows increased down force. Hopefully if I can get more high speed runs I can see the data logging trends and measure the increase in traction.
I did try running the car in reverse. I flipping the front suspension upside down to have the right caster. I found it was stable at low speeds
@@IndeterminateDesign Thank you for the answer. Something about the simulation doesn't look right, the lowest pressure is way over the kick point (where you expect to see it), almost to the middle of the diffuser where the air is slowed down. Usually you expect that on the throat zone to have the high air velocity but from what I am seeing at 5:18 this is not the case. You can search "Aero analysis for F1 2022 based on CFD result" and see the lowest pressure area of an F1 car.
Amazing work with each video !! Which software you used for aerodynamics simulation ? And is the car all in pla ? Would you fear the scenario where you forgot it in the car while the car is in the hot sun which deform it due to high temp inside your car ?
I’m using Simscale for a lot of the “rough” CFD work and I’m using BlueCFD (OpenFOAM for windows) for the fiber meshing. The body of the car is all PLA, the wheels and suspension are all carbon nylon or carbon polycarbonate. I use PLA for my prototypes, and I was in a rush and decided to try to run the PLA wheel and it wasn’t great.
I’m pretty careful to no leave the streamliner in a hot car. It’s actually PLA+ and it has a bit higher melt temp than normal PLA. I have had the ESC make the top of the monocoque soft if I do back to back runs. I may add vents for testing. Eventually the whole thing will be printed in polycarbonate to save weight, but I haven’t gotten to that point yet.
epic
Hopefully constructive opinion:
1. Forget all the aero stuff. You need a controllable reliable car first. Focus on that.
1a. Get a good GPS module to log direction/speed/acceleration/deceleration. (Crashing)
2. Most high-speed R/C cars use a type of foam drive tire. Use that first to get a driveable car, then improve on what they have done.
3. You should be able to calculate the rear tire speed at 100mph. Run that on the bench to measure tire growth and reliability. Save that RPM for later. Log..
5. Examine how fast the rear tires are actually spinning when the GPS says you are at 100mph. Compare to estimates to get tire slippage. A big issue with other R/C cars. Log
5. Measure your ESC power usage at 40,50,60,100~ to create a power/speed profile. This will help you develop your aero package. Log.
6. Take a look at modern Drag Racing bikes to get an idea of what the "profile" of the rear tire should be. They have been doing it for years. This will help driveability. Should be the top focus.
7. Put the car down, and drive 5 feet in the direction you want the car to go. Program the steering to follow data from the GPS to follow that direction. Let the car correct itself instead of you trying to drive. This will need lots of learning/programming as you will want the steering to be less sensitive as the car speed increases. Fun!!
8. Use the GPS or other monitoring device to slowly remove the throttle when slowing down. This should help your braking issue.
9. Use a piece of masking tape on the edges and along the bottom of the car. It will show where the car is hitting the ground and it is easily replaced.
10. Determine ride height at different speeds for the front and rear. Use temporary stick-on winglets to adjust the ride height or downforce in the field for estimated body shape changes.
11. The skis should be taped onto the car's bottom. This allows field adjustment to correct for too high/log ride height in the field. Use shims or tape to change the angle of attack!
12. Outboard skis are da bomb. Possibly replaceable PLA nubs/bumpers. Engineer in a quick swap nose as it's first to "wear out". Bring a couple. That's what F1 guys do.
13. Build the swing arm so you can adjust the driving direction in the field. With more power, you don't know where the car is going.
I could go on, but you need to have some fun also. I think the idea is to build a car that is fast/drivable first, measure/log everything, then start making go-fast parts.
Having lots of shims/tape/spacers etc. in the field is super important to make adjustments while testing. And maybe get a cheap tripod.
thx, John
Thank you for all the really well thought out ideas and feedback!
1. I agree. I think this is the most controllable version of the car I've had so far. I really am tempted to wide the wheel base or try a 4 wheel design to make the car even more stable and easier to control.
-I haven't tried to use the headings from the GPS yet. I think I would most likely need to combine the GPS data with the gyro/accelerometer data with a Kalman filter to have something responsive enough. I actually just found a fatal software flaw where the reading of the GPS data was randomly delaying the microcontroller and stopping all of the stability control systems. I've since switched to using the additional processor on the ESP32 for the GPS and logging, and this has dramatically improved the heading control. I really need multiple dedicated processors so I can separate dependencies.
2. I really want to switch to the foam type tire, but I haven't been able to find the right EPDM foam since I have to custom make my own tires since I can't find the unglued foam tires. It requires a higher durometer foam, and the foam I have tried chunked out at less than 80mph.
3. I admittedly haven't done enough tire testing. The rubber tires grow dramatically even with a homemade belt made of dyneema. The tires self clearance the car's body right now. That's a great idea to create a test rig to run the tire for a prolonged period of time. Do you think I need a roller to exert the force from the road?
5. I just added a rear wheel speed sensor to the car. I've also implemented traction control and antilock braking. Additionally I setup linear interpolation on the throttle and brake, which eases on the acceleration or braking.
5a. Characterizing the motor is something I've been working hard on. The traction control I'm using is rate based, which isn't ideal, but it is used by a lot of drag racers. Ideally, I'd have a motor dyno that allows me to fully model the throttle, RPM, and torque output of the motor. The RC car ESC's are not linear in their torque response. Unfortunately I haven't found an off the shelf solution for logging the motor current that's Arduino compatible. I may be forced to build my own. I am building a voltage logging circuit. I'm using a very small battery which is most likely sagging at higher speeds as well.
6. I have found a semi rounded rear tire helps stability. I think with a wider front width I could run a more square tire and have better traction.
7. I do the heading testing with the gyro and accelerometer and they do pretty well up to a point. I need to do more testing now that I've fixed the software bug. I think after 80mph the car is pretty much being driven by aerodynamics as I have virtually no control over it.
8. I've added the deceleration taper based on wheel speed but haven't tested it. I think a parachute will help to stabilize the car as well, I see this a lot on drag racers when they lose the rear end they pull the chute.
9. I like the masking tape idea, plus it adds a little more wear resistance.
10. I really want to log the ride height. Space is the main issue I have right now. I have some lidar sensors that might be useful if I can get the car more reliable.
11. That's a great point. I need much more adjustability. I was hoping to keep the design simple and just reprint parts, but like you said I need to make adjustments in the field.
12. I am trying some polyurethane skiis for the diffuser edges. It is a surprisingly durable material.
13. I agree on this. I have essentially a "spring rubber" setup for the rear swing arm. I just need to print out several variations and test them out.
Like you said, making the car drivable is the most difficult part. It's hard to find the ideal conditions and perfectly smooth surfaces in which to run the car right now. Again, thanks for the feedback on this. I really enjoy trying to think through all of these possibilities.