Hello, Ratzel I have modified the Tmeasy tire code and used the RK integration method to calculate tire relaxation which is the same method as in pacejka, but it still can not give the right results. So it can be concluded that it is not the reason of integrator. I am very confused now, could you give me more clues about how to solve the problem?
在2022年8月22日星期一 UTC+8 20:09:40<[email protected]> 写道： > Hi Stefan, > > Pacejka uses its own integrator for calculation of tire relaxation, a > possible source of instabilities. Rolling straight forward without crashing > is not good enough for practical use. So don't waste your time with it. > > Btw. even electric motors cannot deliver constant torque with increasing > angular speed, since Power = Torque*omega. It is only possible, as long the > max. power of the engine is not reached. > > stefan schrieb am Montag, 22. August 2022 um 11:24:48 UTC+2: > >> Hi Ratzel, >> Thanks for your answer again. I am doing a pure electrical vehicle >> simulation, the torque applied to the vehicle is constant when the electric >> motor start. The process of constant torque will maintain for a time. >> Can it be concluded that the pacejka model has a different integrator >> from other tire models if it is due to the integrator? Since reasonable >> results can be got from pacejka tire model. >> >> 在2022年8月22日星期一 UTC+8 00:46:23<[email protected]> 写道： >> >>> Hi Stefan, >>> >>> your observation is correct. Actually we know this happens only with >>> handling tires. It is better to make a test with a vehicle with a real >>> powertrain. Applying a const torque means that the accelerating power >>> increases with the speed/angular speed which is not realistic. >>> >>> We had some discussions about this issue, most likely it is related to >>> limits in the integrator. There is some work in progress about it, but it >>> will take a while to get a working solution. >>> >>> >>> stefan schrieb am Sonntag, 21. August 2022 um 13:48:01 UTC+2: >>> >>>> Hello Ratzel, >>>> Thanks for your replying, but I still have some questions about that. >>>> Recently I tested all the tire models in HMMWV vehicle with the conditoin >>>> that a constant motor torque was applied to the vehicle. It is true that >>>> the pacejka tire model is >>>> less robust than others, but it can give reasonable results when the >>>> simulation step size 1e-3 was given. The pictures I attached can explain >>>> what I said(The first picture is for PAC02). The acceleration shouldn't >>>> increase with time under the constant motor torque. And I also found that >>>> acceleration was related to the vehicle speed, and acceleration was not >>>> zero when the vehicle had high speed and engine power was cut off. That is >>>> obviously wrong according to the vehicle dynamics. >>>> I want to know why it happened, and it can be described as a bug to >>>> some extent? or I miss something? >>>> And I also want to solve the problem with your guys and I am checking >>>> the codes of tire parts in chrono. I would be very appreciated if you have >>>> more advice for me. >>>> Thanks again! >>>> [image: PAC021MS.png][image: PAC891MS.png][image: PACEJKA1MS.png] >>>> Stefan >>>> >>>> 在2022年8月21日星期日 UTC+8 16:28:13<[email protected]> 写道： >>>> >>>>> Hi Stefan, >>>>> >>>>> the Pacejka tire model in chrono is just outdated. While it is nearly >>>>> feature complete, it lacks stability. Another problem is the data input >>>>> routines. It pretends to read tiem orbit files (Adams/Car), but expects >>>>> the >>>>> data entries to be in a certain order. It is very likely that nonsense >>>>> input is read from a valid Adams/Car tire file, as long as you don't >>>>> reorder the data lines, I guess nobody is eager to do that. >>>>> >>>>> If you want to use a Pacejka tire model in chrono, use Pac02. Pac02 >>>>> hasn't all features possible, but it works and doesn't crash chrono >>>>> vehicle >>>>> simulations. For a benchmark project we needed a Pacejka89 (Adams/Car) >>>>> model. This model has a very clear implementation structure, this >>>>> structure >>>>> was the also taken for Pac02. The slip definitons differ of course, as do >>>>> the formulae. >>>>> >>>>> Pac02 and Pac89 are comparable in simulation accuracy. Pac02 can be >>>>> easily rescaled to different tire load levels due to its skaling factors, >>>>> so it is more flexible to use. >>>>> >>>>> stefan schrieb am Dienstag, 16. August 2022 um 10:43:57 UTC+2: >>>>> >>>>>> Hello, >>>>>> >>>>>> The pacjka tire models expressed in the manual >>>>>> <https://api.projectchrono.org/wheeled_tire.html> are Pac89 and >>>>>> Pac02, But I found pacjak tire model in HMMWV vehicle, and the >>>>>> simulation >>>>>> results derived from Pac02 and Pacjka model are obviously different, I >>>>>> don't know which model is better for vehicle simulation and which >>>>>> derived >>>>>> simulation result is closer to the real one. Otherwise, could you tell >>>>>> me >>>>>> the difference between Pacjka model and Pac02 model from the >>>>>> implementation >>>>>> principle. >>>>>> >>>>> -- You received this message because you are subscribed to the Google Groups "ProjectChrono" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/projectchrono/f056c74b-f9b6-4a6a-85db-8a30596d3bb3n%40googlegroups.com.
