Crank/valve train model 2

[SMUG]

All

What are the industry standard softwares for crank/valve train modelling for automotive engines?

Has anybody evaluated these tools?

What are the limitations on these??

 

[crysta1c1ear]

http://www.profblairandassociates.com/GPB_Products_Main.html

You could use your requirements and look at the above site to create a checklist of what you might want to achieve and then measure any products by seeing how much of your checklist is offered by particular products.

 

[ivymike]

Some OEMs have their own software for this, but there are commercially available packages too. A couple that come to mind are ADAMS Engine and Ricardo Valdyn. I believe that there is also a set of powertrain analysis tools for the general simulation code EASY-5.

 

[SMUG]

I have seen another one called AVL Exite. Does anybody use these tools in the forum?

 

[SomptingGuy]

It kind of depends what you mean by "crank/valve train modelling". If it's coupled crank/block FEA then you're talking AVL Exite/Ricardo ENGDYN. But if it's lumped parameter stuff, I'm not sure anyone else does the equivalent of Ricardo VALDYN.

 

[GregLocock]

valve/camshaft/FEAD dynamics can be done in ADAMS/Engine. I have been asked to do that, seems like a good way of looking at FEADs in particualr.

I'm not sure that the roundy roundy uppy downy bits (cranks and pistons) really gain much from ADAMS analysis although I suppose you could examine piston slap and bearing loads.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[SomptingGuy]

In a previous position several years ago my job was to build models of the "roundy roundy uppy downy bits" using ADAMS. Although it generally worked (could predict torsional speed variations and other low-frequency things quite accurately), it did seem like a sledgehammer to crack a nut (solving 1D systems with a 3D code).

I was also asked to model FEADs to predict gear rattle in the timing drive of a big diesel. That was even less suited to ADAMS than the "roundy roundy uppy downy bits" - a highly stiff 1D system.

This is not intended to be a pop at ADAMS. I liked using it and built many interesting models. I just got the impression that we had an expensive tool looking for problems to solve to justify its annual lease. And many of those problems could have been solved quicker, cheaper and more accurately using other tools.

 

[SMUG]

Is it good idea to look at cranktrain and valve train in assembly or looking at in isolation will produce the same results.What does FEAD dynamics produce in ADAMS.

Does it mean you cannot do coupled (crank/block)analysis within ADAMS.

 

[SMUG]

I am looking at increasing the displacement of an existing petrol engine with the minimum number of parts modification, obviously want to get confidence from CAE front. Any ideas on which components are the most likely candidate to be replaced or modified and what are the economical codes to go for.

 

[SomptingGuy]

Does it mean you cannot do coupled (crank/block)analysis within ADAMS.

Given enough time and motivation you can do anything, appropriate or not. ADAMS is a lumped parameter modelling system and models those types of systems best. FEA is inherently a distributed parameter modelling system. While there are methods of including reduced FEA models in lumped parameter models, it's always a shoe-horning job.

Take a look at MSC's promo material:

http://www.mscsoftware.com/support/library/conf/adams/euro/2000/FEV_Powertrain_Dynamics.pdf

 

[facty]

The site mentioned above by Crysta1c1ear only applies to valvetrains, not cranks:

http://www.profblairandassociates.com/GPB_Products_Main.html

However, on their website they include several very interesting articles they published in "Race Engine Technology" on the Cam and Valvetrain Design process and how this is best applied using their software.

These are not tech articles like SAE etc, but are instead very intuative guides for the Cam/Valvetrain designer.

 

[SMUG]

SomtingGuy thanks for your valueable links. Reading at the artical explains that hydrodynamic forces will increase in mycase and the most likely modified candidate for me will be bearings,crank and maybe piston.I suppose valvetrain would not be effected.Any ideas.

Thanks for the the clarification facty.

 

[facty]

Woops - sorry for the spelling - should be "intuitive"!!

The articles that I downloaded from their site (mentioned in my initial message) are in pdf format and are definitely worth the read if you're involved in Cam/valvetrain design.

 

[SMUG]

Facty, What are your thoughts in most likely components which will need modifications?

 

[facty]

SMUG, the answer to your question requires much more detail re the purpose of your design problem - why are you increasing the displacement? - more performance etc? What is the spec/use of the existing engine?

I agree with your approach on wanting to investigate the design boundaries using CAE tools. For this type of investigation I would use a 1D engine performance simulation tool (Optimum Power, Gamma or Ricardo) to analyse the potential performance gains from changing geometry and valve timings/lifts etc. Having obtained an optimum design I would then check that it was mechanically feasible by using a tool such as 4stHEAD that is described on the Blair website (others would be GT-Vtrain, Valdyn etc).

Finally the mechanics of the crank could be analysed either separately or linked, depending on what CAE tools I had at my disposal.

I have used Optimum Power's Virtual Engines 1D software in the design arena before and have experienced reasonable success with it. The key to using any 1D software is to appreciate and realise their frailties.

Regarding the Cam Profile Design and Valvetrain Analysis tools out there, I have recently heard very good reports on the Blair software and saw the Presentation he gave at the motorsport Symposium in London back in November. However Gamma and Ricardo also have tools for this application.

In high performance engines I have seen significant gains from investigating the coupled effects of the torsional vibrations in the crank and cam and how this will affect the ensuing engine performance.

I could go on for ages, but the bottom line regarding CAE tools is your own specific design applications, resources and future plans.

All CAE tools are very expensive and can only really repay their investment if they are being constantly utilised. Some may even be the proverbial "hammer to crack an egg" so it is important to know exactly what they are all capable of doing, their strengths and weaknesses, although the latter 2 may only be apparent after you start using them in anger!

Most people tend to play safe so the perceived "bigger" CAE suppliers get bigger at the expense of those that may suit your purposes better. The plus is that most CAE tools have yearly licences, so you can chop and change if you are fortunate enough to have the time to do so.

I think I've gone on for long enough!

 

[GregLocock]

If you want a free 1d engine code wander over to

http://www.lesoft.co.uk

I've only played with it, not having an engine project on the go, but it seems to be a useful piece of work.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[facty]

Greg's right - I forgot about the Lotus 1D software. The "free" download has a drawback - it can only model a single cylinder engine and therefore you will not be able to investigate the intake and exhaust tuning effects of a multi cylinder engine. However, it would give you a feel for the whole 1D modelling scene and you could factor the performance results in proportion to the number of cylinders your engine has (to get ballpark figures!).

Alternatively the other 1D code suppliers may give you demo versions if requested.

 

[SMUG]

Thanks facty for the in depth info. This is a effectively a learning exercise where I want to improve the performance
an existing four stroke engine 1600cc by altering the geometry while keeping the number of carry over parts as minimum as possible .

Hence as you suggested 1d tool will be used for performance, but at the same time the boundary conditions i.e forces,pressure etc on components will also change and I want to see the boundary conditions after the mods to make sure the structure is sort of ok. What tools would be ideal for this type of work.

 

[SomptingGuy]

How many of these new engines are you going to be making?

 

[SMUG]

One off. As I said earlier this is purely a learning exercise.I have intentions to use k-series engine as its block and lining is generic for higher displacement engine. Any different ideas are welcome.