[Jinjolee]

I am currently working on a project that needs to perform a CFD to simulate the drag while a cyclist is driving along a track loop. What I have been doing is simulating the drag of the cyclist driving along a straight path, by setting the inlet velocity and set the cyclist geometry static, like a wind tunnel experiment. I believe this is most of the researches of studies did. However, one of the important part of my project is to simulate the drag while undergoing a turning motion at a corner. To perform this task, I came up with two ideas.

One is to give motion to the geometry inside the fluid domain. However, this would result in a huge fluid domain and some dynamic mesh setting, and I doubt that the computational power of my PC would not be able to perform it.

Another method is to use the similar approach of what I did for the straight path. Let the geometry be static and change the inlet velocity instead. However, I do not know how to model this. When the cyclist is turning, says turning left, then at a instantaneous moment, the air on the left should be somehow be pushed by the cyclist(I am not sure if I describe it correctly). But how could I obtain these values? Without knowing the velocity field under the turning motion, it is hard to set the input values for the inlet velocity. I have searched the internet, and it seems most of the studies did not consider such kind of motion. I am quite helpless at this point.

Could anyone give me some idea or suggestion? Or just leave some related reference. I appreciate your kindness and your help, thanks a lot!

[FMDenaro]

Originally Posted by Jinjolee

I am currently working on a project that needs to perform a CFD to simulate the drag while a cyclist is driving along a track loop. What I have been doing is simulating the drag of the cyclist driving along a straight path, by setting the inlet velocity and set the cyclist geometry static, like a wind tunnel experiment. I believe this is most of the researches of studies did. However, one of the important part of my project is to simulate the drag while undergoing a turning motion at a corner. To perform this task, I came up with two ideas.

One is to give motion to the geometry inside the fluid domain. However, this would result in a huge fluid domain and some dynamic mesh setting, and I doubt that the computational power of my PC would not be able to perform it.

Another method is to use the similar approach of what I did for the straight path. Let the geometry be static and change the inlet velocity instead. However, I do not know how to model this. When the cyclist is turning, says turning left, then at a instantaneous moment, the air on the left should be somehow be pushed by the cyclist(I am not sure if I describe it correctly). But how could I obtain these values? Without knowing the velocity field under the turning motion, it is hard to set the input values for the inlet velocity. I have searched the internet, and it seems most of the studies did not consider such kind of motion. I am quite helpless at this point.

Could anyone give me some idea or suggestion? Or just leave some related reference. I appreciate your kindness and your help, thanks a lot!

Well, I don't know if you can get some results searching references in internet...Recently a very massive computation of peloton has been done but using a static configuration.
On the other hand, if you look at a track cycling, it is deigned in such a way that the cyclist is almost normal to the wall in any point. Thus, are you sure that is relevant the movement along the track? And at this point, I wonder if would be much more relevant to take into account the real legs movement...

[Jinjolee]

Originally Posted by FMDenaro

Well, I don't know if you can get some results searching references in internet...Recently a very massive computation of peloton has been done but using a static configuration.
On the other hand, if you look at a track cycling, it is deigned in such a way that the cyclist is almost normal to the wall in any point. Thus, are you sure that is relevant the movement along the track? And at this point, I wonder if would be much more relevant to take into account the real legs movement...

Thanks for your reply. For the computation of peloton you have mentioned, what is the computation about? I would like to know about that.

And do you mean that since the cyclist is almost normal to the wall, the motion of cyclist on that semi-circular turning path is nearly the same as on a straight path? I am sorry that my English is bad and may not be fully understand what you said. Sorry about that.

[arjun]

Originally Posted by Jinjolee

I am currently working on a project that needs to perform a CFD to simulate the drag while a cyclist is driving along a track loop. What I have been doing is simulating the drag of the cyclist driving along a straight path, by setting the inlet velocity and set the cyclist geometry static, like a wind tunnel experiment. I believe this is most of the researches of studies did. However, one of the important part of my project is to simulate the drag while undergoing a turning motion at a corner. To perform this task, I came up with two ideas.

One is to give motion to the geometry inside the fluid domain. However, this would result in a huge fluid domain and some dynamic mesh setting, and I doubt that the computational power of my PC would not be able to perform it.

Another method is to use the similar approach of what I did for the straight path. Let the geometry be static and change the inlet velocity instead. However, I do not know how to model this. When the cyclist is turning, says turning left, then at a instantaneous moment, the air on the left should be somehow be pushed by the cyclist(I am not sure if I describe it correctly). But how could I obtain these values? Without knowing the velocity field under the turning motion, it is hard to set the input values for the inlet velocity. I have searched the internet, and it seems most of the studies did not consider such kind of motion. I am quite helpless at this point.

Could anyone give me some idea or suggestion? Or just leave some related reference. I appreciate your kindness and your help, thanks a lot!

It all depends on the computational resources you have. If you have resources it could be done i guess.

[FMDenaro]

Originally Posted by Jinjolee

Thanks for your reply. For the computation of peloton you have mentioned, what is the computation about? I would like to know about that.

And do you mean that since the cyclist is almost normal to the wall, the motion of cyclist on that semi-circular turning path is nearly the same as on a straight path? I am sorry that my English is bad and may not be fully understand what you said. Sorry about that.

Have a look at this paper https://www.sciencedirect.com/scienc...67610518303751


and this topic Wind tunnel model of an animated cyclist

[Jinjolee]

Originally Posted by arjun

It all depends on the computational resources you have. If you have resources it could be done i guess.

If I have adequate resources, does that mean I should go for the option that directly move the geometry inside the fluid domain? Thanks.

[Jinjolee]

Originally Posted by FMDenaro

Have a look at this paper https://www.sciencedirect.com/scienc...67610518303751


and this topic Wind tunnel model of an animated cyclist

Thanks for your materials, I will take a look into them. Actually I am also doing the track cycling simulation, but my supervisor seems not satisfy the work I have done, because she thinks the slope and the curvature are the characteristic of the velodrome, and I should focus more on making use of these characteristic onto my CFD simulation. meaning that I should not just perform the straight track simulation. However, it seems that not much people would take those into account in CFD, plus I do not have much knowledge in CFD.

[arjun]

Originally Posted by Jinjolee

If I have adequate resources, does that mean I should go for the option that directly move the geometry inside the fluid domain? Thanks.

The link that Prof Denaro posted you will see that in that thread i posted a link to simulation with moving tyres on a bike.



I used immersed boundary to move those tyres. This was done to show how immersed boundary is solved in the solver.



I used 5 million cell mesh along with 32 process to run the calculation.



Now you can move the whole bike with custom motion but you will need finer mesh to treat full bike. This is what i meant.

[billsmartin]

Hello Everyone. Everybody knows that oil prices are increasing day by day as well as vehicle prices are also increasing direct proportionally. ZSo here is a solution that we have to get electric bicycles. Just click on the link and you will get what you will want https://rideonelectric.com/electric-bikes/
stay blessed

[vivjk94]

Originally Posted by Jinjolee

I am currently working on a project that needs to perform a CFD to simulate the drag while a cyclist is driving along a track loop. What I have been doing is simulating the drag of the cyclist driving along a straight path, by setting the inlet velocity and set the cyclist geometry static, like a wind tunnel experiment. I believe this is most of the researches of studies did. However, one of the important part of my project is to simulate the drag while undergoing a turning motion at a corner. To perform this task, I came up with two ideas.

One is to give motion to the geometry inside the fluid domain. However, this would result in a huge fluid domain and some dynamic mesh setting, and I doubt that the computational power of my PC would not be able to perform it.

Another method is to use the similar approach of what I did for the straight path. Let the geometry be static and change the inlet velocity instead. However, I do not know how to model this. When the cyclist is turning, says turning left, then at a instantaneous moment, the air on the left should be somehow be pushed by the cyclist(I am not sure if I describe it correctly). But how could I obtain these values? Without knowing the velocity field under the turning motion, it is hard to set the input values for the inlet velocity. I have searched the internet, and it seems most of the studies did not consider such kind of motion. I am quite helpless at this point.

Could anyone give me some idea or suggestion? Or just leave some related reference. I appreciate your kindness and your help, thanks a lot!

In order to do something this unique, you need to find a journal paper who has done something like this and use it to do your analysis or better yet, find a bicycle, attach a few sensors and perform an experimental analysis yourself.

[aerosayan]

@OP

You won't be able to perfectly model everything. Do basic tests, and then do more complex tests, and then make guesses, or read other research papers, or do real-life experiments on what can't be tested.

1. Read prior research, and find the simulations they're doing, and the results they got. If tour-de-france competitors aren't worried about doing convoluted tests, your project won't need them either. Try to mimic their procedures, as they have more incentive to get accurate results, and have more money to spend on their R&D.

2. Take different configurations of the rider, and wind flow, and do static tests.

3. As FMDenaro said, the movement of the legs might introduce a lot of drag, and I agree. It might not be feasible for you to do a dynamic test, if your PC specs are low. Try to read other research papers to find data, and try to make a guess.