Solidworks transient simulation

[beshr]

Hello everyone

My name is BSHER and I am a member of JUST racing team. I am writing this to seek some answers to some problems I faced the other day while preparing our intake system for transient flow simulation by solidworks.
I searched over the internet on how to perform such analysis, eventually I managed to learn how to it. After running the simulation and setting a sequence for the intake strokes for the runners, I got some unreasonable results as it was assumed that the runners were opened to the atmosphere and acted as inlets during the intervals where they should be closed. So please, if there is anyone who encountered such a setback can help me?
Thank you for time.
P.S.: I used a polynomial function to describe the pressure drop throughout the intake stoke.

[Kareem]

Did you checked "Time dependent" option in the wizard tool?

and please can you discuss how did you managed to get a relation between time intervals and pressure?

[Danschwind]

Can't help you with the software since I have never used SW Flow Simulation. But considering the not-so-great Static Simulation by SW, I would recommend using a better sofware like ANSYS or else.


But on a more general question, what is your goal with this simulation? What to you want to optimize? Do you already have a car that can move by it's own?

[apalrd]

Did you analyze the rest of the manifold first?

I can think of a few better uses for simulation time/effort than looking at transient flows in a manifold.

A transient flows simulation is 3-d. In general, you add dimensions as they are actually needed, and start from 0 dimensions.

You will get nearly the same performance by calculating the required runner lengths (INTAKE AND EXHAUST!!) in a spreadsheet, using good engineering judgement in preventing sharp turns and tapers around the restrictor.


Overall, the total volume between the restrictor and the valves is far more important than the shape of that volume. You can find the required volume with 1-D simulation or an exceptional spreadsheet.


If you did 1-D simulation of the engine as a whole, you would be able to look at the valve flow vs crank angle in your chosen simulator and copy that into the 3-D tool.

[Danschwind]

I agree with Andrew, that is what I was going to suggest after you answered my question :D

I strongly advise you to focus your simulation efforts on a reliable and solid 1-D simulation, using either Ricardo WAVE or GT-Power (or any other good software), since the overall benefits of it are far greater than those of a 3-D simulation in that case. You can use the 3-D simulation later to improve your design, but do not focus solely on it.

[beshr]

Dear Kareem,
Thank you for your response. The answer to your first question is: Yes, of course.
The answer to your second question is:
1- Determine certain rpm you want (let's say 3000 rpm),this means there are 1500 complete cycles in 1 minute, this means that one cycle needs (60/1500) seconds to be completed.
2- From your engine's catalog or from the internet, determine the intake stroke interval in degrees (in my case 284 degrees).
3- To find the pressure distribution over the 720 degrees of the crank angle, I assumed that the function of the pressure drop during the intake interval is polynomial (as the cam profile).
4- Find the expression.
5- The rest of the interval is assumed to be constant (closed).
6- Repeat it for each cylinder.
I hope that helps.

[beshr]

Dear Daniel,
Thank you for your response. We are performing this kind of study in order to produce a high quality intake system for future designs (class 2). We want to study and optimize each parameter of the intake system and its effect on the mass flow rate and its distribution.
Regarding our car, we still work on it.
I tried to use the Ansys earlier today instead of solidworks. I followed the instructions of a detailed youtube video, but I also get unreasonable results as it acted as if there is only one runner and stayed that way throughout the entire run.
So, please if you have some advice, I would appreciate it.
The link of the youtube video:
https://www.youtube.com/watch?v=xsZleJdmGp8

[beshr]

Dear Andrew and Daniel,
We are doing this study as a research and also we have untraditional configuration for the intake system in mind that differs than the configurations that are used.
May I ask you about Ricardo wave? Is it accurate? Is it easy to learn? And most importantly does it accept complicated 3d models (from solidworks or creo)? Or just it deals with straight pipes and elbows?

[Danschwind]

Beshr,

I am not too much into WAVE, but since it has most of the capabilities of GT-Power, I guess yes, you can use complex 3-D models in it. The problem comes with the discretization of such geometries, since it will be regarded as being made of multiple equal-volume cubes.

This "untraditional configuration" is just an odd geometry or there is more to it?

The models in 1-D simulation is as accurate as you dedicate time to build it. You need of course some sort of validation for your model, either being with an in-cylinder pressure sensor or comparing the expected torque curve with the real one you can get from a dyno. Of course there will be many differences in both curves, but if your valleys and peaks are kinda in the same RPM range you have a close model.
If you dedicate enough time and efforts, you can make it very very accurate. I've seen a model of a real engine in GT-Power having a average error of around 5%, it was pretty good.

It is not hard to learn. What is hard sometimes is to measure the entire engine to create the model, but if you already have the cam-profile and such, it will be way easier.

Are you inputing acoustic effects in your runner simulation?

[beshr]

Daniel,

You can say it has an odd geometry (of course for specific reasons).

Regarding the last question, if you mean (Ansys), I just selected "transient" and define functions and expressions from an excel sheet (as mentioned before) as the outlet boundary conditions then I started the run.

If this isn't what you are asking about, please explain further more.

Thanks

[Danschwind]

Beshr,

I'll try to answer you with a question: how have you calculated/defined your runner lenght and diameter?

[beshr]

Daniel,
I know that I have to determine the length of the runner such that it assures that the compression wave is located behind the valve just before it opens. But I didn't determine a specific value as the runner length is one of the parameters that I want to study its effect on the system.
I thought that the acoustic effects will be included in the simulation as I determine the boundary conditions at the outlets. Please, correct me if I am wrong and how to include it in the simulation?
Thanks.

[Danschwind]

I can't give you a precise answer to this question because I've never used a 3-D simulation to design my intake runner and plenum.

But, again: I highly advise you to ditch the 3-D simulation for now and start working on a 1-D simulation ASAP. All your questions will be answered with it.

----

If your intake valve start opening when you piston is almost at TDC, will you want the peak of your reflection wave pressure to still happen at this moment?

[beshr]

First of all, I appreciate your advice and I think I will take it into consideration. Second of all, my basic answer would be yes, although I sense there is a point you want to make behind this question.

[Danschwind]

If the peak happens before the valve opens, you would be losing some of its momentum to reduce the backflow during overlap, right? When the overlap happens, usually covering a small range around TDC, your goal is to reduce backflow as much as you can, so that wasted momentum, that hit the back of the intake valve before it was losing its fight to the exhaust valve would make some difference, right? :)
----------------------


Short story, when I first joined my team back in 2011, there was a guy who was really really amazing with Fluid Mechanics designing the plenum. He and 2 other guys who were already seniors in the university spent 4 to 5 months just running some SW analysis of the geometry and trying their best to get to a point where the simulated flow rates between each cylinder to be as equal as possible. The plenum turn out to be a pain to fabricate, and when put to use, its performance was poorer than all cylinder-shaped plenums where the guys actually spent time calculating the runner lengths of both intake and exhaust to match the desired performance at the desired rpm. So, advising again: work your way on intake and exhaust lengths, use a simple plenum shape and after you get a satisfactory performance, you can do those tweaks at the geometry :)

[beshr]

Daniel,
Thank you so much for the advice and the information. Now, things make sense to me. I thought that the purpose of the compression waves were to act as a boost for the intake
charge. I didn't look at it this way.

[Danschwind]

The absolute orgasm for an engine guy is to have a forward pressure wave with peaks both at the overlap and when the valve is almost closed (you can think of that as a little "punch" on that last minute where not that much air is being sucked into the cylinder because the piston velocity is very low).

The word "boost" can be used, yes. But that "boost" during overlap is more like a "don't let the backflow ruin my performance" than a proper cleaning of the cylinder.

Dunno if Ricardo WAVE has a plot like this, but GT-Power has a "forward waves (or something like that)" that is very very very nice to analyse. You will spot a increase in pressure at the back of the valve during overlap, but the "back" wave (going from the cylinder to the intake) is generally higher in pressure than the "forward" one. Only under some circumstances you are able to have a bigger "forward" wave, and when that happens = POWAH!

[beshr]

Thank you very much for the information. It was very helpful discussion :)

[Danschwind]

I'm glad to help. If any engine guru want to add something, it would be nice :D