Hello, we are searching for any information on the web about the Ricardo Software named WAVE. This sofware is a 1D engine simulator.

Thank you for your Help.

CED

Have you looked at Ricardo's homepage? If not, go to (se menu above) Resources/Homepage Database/Companies. There is a link to Ricardo there.

Ricardo is a small company and the guys who developed the software don't have credentials just like the case of software Star-CD. Wave is a software for small people (low level engineers) or the beginners in compressible flow. Kiva is comparatively better hydrodynamics software developed in National Laboratories by people with good academic credentials but still more work needs to be done. Fluent is comparatively better than Star-CD because Fluent gives better results than Star-CD. Star-CD may be easier to converge than Fluent but it doesn't mean the results are more satisfactory. So if you can't get hold of any software developed in National Laboratories or NASA then I would recommend you to stick to Fluent. Something is better than nothing. Special softwares are developed to serve special purposes which have great significant importance in science and engineering but we may not have access to it. We have to learn to live with this limitations. We can't do anything about it.

Eh, you're comparing apples and oranges. Wave is a highly specialised 1D gad-dynamics code used solely to design engines. It is not a CFD code and it can of course not be replaced by a general purpose CFD code, like the ones you are mentioning. In fact I even think that Wave can be coupled to, for example, STAR-CD, if you want to model some components using CFD.

I have no idea what made you compare WAVE with general purpose CFD codes like FLUENT or STAR-CD. These are two completely different groups of products which have little or no similarity at all. I am a WAVE user and I object to the saying WAVE is for 'small people(!)'. Have some respect for people who do not follow the same methodology as you do.

WAVE is a 1D gas dynamics software used mainly to model internal combustion engine performance and acoustics. It is a widely accepted software for powertrain analysis not only in the US but worldwide. It is true that 1D formulation has a lot of shortcomings over 3D formulation but this does not mean it cannot deliver valuable information about the flow in general.

If you do want to compare WAVE do that with a similar product such as GT-POWER which is another powertrain analysis code based on 1D assumption. Both these codes have interface with 3D codes such that the 3D code output may be used as input to the 1D model and vice-versa. WAVE is connected to VECTIS and GT-POWER is connected to STAR-CD.

Both WAVE and GT-POWER have a nice web site. Please visit www.ricardo.com and www.gtisoft.com for detailed information.

Didn't the companys designed their diesel engines using WAVE, STAR_CD or whatever crab and got penalty of one billion dollar for producing engines with emissions problems? Then how can these softwares be reliable! You are not in USA and you are not quite aware of what is happening. GM already started spending about one billion dollar per year until 2004 to develop engines with less tailpipe emissions, better performance and meet fuel economy goal. Daimler-Chrysler and Ford Motors are also spending billions of dollars for research to come up with solutions to meet future stringent emissions goal. We are seeing with our own eyes the consequences of using wrong softwares. This is not a child's play anymore. So now you can guess where small companies who developed STAR-CD or WAVE stands!

Usually I would let baseless statements go unanswered, since this is a forum for CFD and related issues. However, I feel it is necessary to respond to this thread.

(1) the size of a company is not an indication of the quality of its output (design, software, analysis, etc.). In fact, I can argue the exact opposite -> a small company survives only by being at the forefront of technology and by working hard. In contrast, many (not all) people who work for giants spend most of their time in the halls of the buildings with their coffee mugs in their hands and talking BS. I have worked for the smallest to the largest of them - I know!

(2) I have not used any commercial codes so I am not supporting one vs another. But the assumption that you will have all your problems solved using the most complicated 3D code does not correspond to reality. See the following:

I understand that WAVE is an acoustics code. Flow in engines is for the most part incompressible - or low Mach number compressible. To the second order in the Mach number you can obtain accurate solutions by the assumption of incompressibility. What does this mean? This means that the effect of pressure is felt everywhere instantaneously, because (again up to the leading order of the Mach number) pressure variation depends on time and not space. By extension, it becomes obvious then that, for the bulk of the flow, acoustics in the engine is 1D. The 3D effects are secondary in nature (you have much more serious issues to contend with than the three dimensionality of the acoustics in engines).

So, what does this tell you about a person who uses a 1D code for engine acoustics analysis vs. a full blown CFD? I'd say, if the code is used in the proper range, the first guy knows what (s)he is doing. I cannot be impressed by the CFD user who believes in brute force and expects the colored results to somehow magically help him understand the flow physics!

As for the tremendous loss engine companies are suffering from -> the conclusion that it is the fault of the CFD codes (or developers) and not the users is equivalent to blaming the gun itself for the action of the murderer (sorry I could not think of a better analogy)! Somehow you seem to be under the impression that people who work for these big engine companies know what they're doing (know their physics, can interpret data and not just show colors to dazzle their bosses, and more importantly can provide better solutions) Absolutely wrong: I have seen an internal report by an engineer for an american auto company who did a parameteric study using CFD, varying Reynolds number and viscosity as two parameters of interest! They also concluded that viscosity is as important a factor as the Reynolds number! (That is if you fixed the Reynolds number and changed viscosity you would get different results, blah blah) What a pile of garbage - not even the boss of this engineer (with a PhD) objected to the absurdity of the choice of parameters and the conclusion. It tells you that this guy doesn't know the very basics of fluid mechanics and you expect reduced emissions coming out of this particular center of excellence?

Let's be a little more prudent in our observations and conclusions of the reality.

Adrin Gharakhani

National Loboratories and leading universities are putting their heads together to come up with solutions to meet future stringent emissions problems in USA. WAVE or STAR-CD softwares are like mole-hills compared to mountain. I am repeating again that this is not a child's play anymore. You are probably trying to build a mole-hill with sand that has not chance of survival.

In defence of both companies that you are describing as 'small fry'. Admittedly I have worked for/with both of them so perhaps I am biased.

One, Ricardo, has been in the internal combustion engine development business since approximately 1920, well before Los Alamos was a twinkle in Robert Oppenheimer's eye. Now 70 years developmental experience may not mean anything to you but in the industry they are well respected. They probably, as an organisation, know a little bit more about the workings of the internal combustion engine than the average national lab.

As for STAR-CD, CD its parent has developed Speed which was funded by the European auto industry again a not insignificant level of experience. As for them not having credentials, surely coming from Imperial College for finite volume CFD codes is the credential.

I would be interested to know what 'better results' means ? Are you saying that on the same grid using the same boundary conditions, similar if not exactly the same numerical scheme and with the level of convergence equivalent (taking into account the different definitions used) you achieved fundamentally different results? If so, I would have to say that one of the codes had a bug or that the comparison is flawed in some other way. The various commercial codes have their many strengths and weaknesses but should produce comparable results if compared correctly.

Finally by the standards of GM/Ford .... Fluent (Aavid) is a tiddler. Does this mean that GM should only be using software developed at their own research labs. Competency and size are not one and the same thing.

(1). I don't know what to say about the issues here. (2). But I think, I need to point out something. When the name of the code is mentioned, when the name of the company is mentioned, somehow, it becomes more important than the CFD itself. (3). I think, in the IC engine design, flow through the inlet manifolds and the exhaust manifolds can have significant effect on the flow through engine, especially the multi-cylinder engine. So the length has to be computed based on the wave equations. There are books on this issue in great details. (4). There are also other important factors which affect the design of the engine itself, such as the fuel injection , the mixing, the way the flow entering the inlet port, the flow motion through the valve, the valve timing, the valve timing control as function of speed, the induced flow in cylinder, the flow rotation, the tumbling, the spark plug design, the heat range and heat transfer, the cylinder head design, the effect of turbulence on the combustion, the pre-ignition, the lean-burn design, the design of catalytic converter, the control of the catalytic converter,..... (5). I think, the traditional 1-D tools is o.k., but it can provide the overall performance model, with a lot of empirical data. The Star-CD although has the tradition from the Imperial College, is still far from providing the answers to the above design problems mentioned above. The nice thing about the Fluent is its unstructured mesh and refinement capability. But that alone, is not going to solve the problems. (6). I would say that it is going to take a lot of hard work to make progress in this field. This is because it is hard to make detailed measurement inside and outside the engine. Hopefully, the billion dollar project will first make the detailed flow field available first, and thus provide the information needed for creative new aproaches or design. Without such detailed information, trial-and-error approach is not going to solve these problems effectively. (7). It is not the codes which keep the company alive, I have seen 40 years old company disappearing, not to mention a company with over a hundred years of history, which also has disappeared recently. There is no question about the ability of these companies to design the product, the issue is can they design the product to meet the future standards and be able to say why at the same time. (8). Naturally, the key issue is in the fluid dynamics. And I don't think the traditional 1-D code will add any new information to the understanding of fluid dynamics, not even the current version of commercial CFD codes.

Hi all,

We are currently comparing StarCd and Fluent for unstructered grids and I am extremely interested in your statement that Fluent is better than Star. On what do you base this statement? Which benchmarkcase have you simulated?

(1). We are happy to learn that you and your company are taking the serious approach. (2). I can only share with you my experience in the industrial use of CFD. (3). The mesh independent solution is never resolved in industries, because it is not practical. It takes too long. (4). The solution convergence issue is never resolved, because in industries, they don't have a lot of money to spend. The solutions used in design are most likely non-converged solution. (If a case requires 80 hours to reach convergence, by just stopping the calculation at 40 hours will move the project ahead of time and there will be no problems with milestones to meet.) (5). The mesh is routinely optimized to give better solutions. If you can't refine the mesh to improve the solution, you find the mesh which can give you the better solution. (6). These are fairly common in industries, because the user is not the developer. So, he is not emotionally attached to the codes or solutions. And in many cases, he is not the one who is using the solutions. So, he does not care. (7). So, in the industrial society which is always looking for ways to cut costs, time, nothing can stop it until something bad happens. In other words, the industrial society prevents the proper use of CFD in industries. (The code testing at the user's site is already too late, from my point of view. The code should be in perfect condition to solve the user's problem when it arrives at the user's site. Otherwise, it should be returned. This is still my personal opinion. Believe me, it can be done.)

Convective Heat Transfer from Hydrocarbon Flames

In internal combustion engines, a primary cause of low thermal efficiency is the heat transfer from the combustion gases to the walls of the cylinder during combustion. This research is directed toward two methods for controlling the convective heat transfer from such flames. The first method uses an electric field to force charged particles generated within the flame away from the nearby wall. In moving away from the wall, the particles thicken the wall boundary layer, thus reducing the wall heat transfer. The second method of control is to produce a tangential velocity over the wall surface with an external jet of air, thus moving the flame away from the wall. The objective of each of these studies is to reduce the wall heat transfer. Incorporating these techniques within the combustion chamber of internal combustion engines will significantly improve the overall efficiency of these engines.

Increasing the strength of swirl generation by tangential jets can reduce screeching noise in an engine. Can 'WAVE' software handle this kind of convective heat transfer and acoustics problems? I guess not! I have already mentioned earliar that I am a novice in CFD and am trying to learn.

Fluent uses finite element method which is comparatively more matured method. Fluent usally is harder to converge for a given flow of interest but if you can converge it it will give you better results than STAR-CD. Try it yourself and let us know the results. Structured-based meshes are simpler and easiar to converge than unstructured meshes. Nowadays, even NASA is more interested to develop softwares using structured based simpler method. I really don't trust commercial CFD softwares that much but if I badly need to use a commercial software then I would stick to Fluent.

Both Fluent and Star use Finite-Volume on unstructured grids.

Yes! Both Fluent and STAR-CD use the finite volume approach and, in fact, STAR-CD was the first successful commercial code using unstructured mesh topology with finite volume approach.

CFD covers wide range of topics and CFD results strongly depending on the PHYSICAL modelling of things like turbulence and combustion - both topics have been attracting the attention of many leading scientists. If the physical models do not reflect the reality accurately, one should not expect the ACCURATE CFD results - I think many experienced CFD engineers are aware of this.

WAVE and GT-POWER are very useful specialised software and 3D CFD codes and these codes can compliment each other nicely.

I am the business development manager for Ricardo Software. Please find the information below in response to some of the questions/points raised in the messages posted.

Ricardo is an engine engineering consulting company with operating offices all around the world including 3 in the US. We now have 1300 employees, the majority being engineers holding advanced degrees. We have been working on IC engines for over 85 years, so we are very experienced in this field. Among our contributions to the development of the IC engine that are part of our everyday life I can quote the definition of the Octane rating scale for fuel (the number you see at the gas station). We have developed a line of specialized CAE software tools (including WAVE) for the design of IC engines and specific engine components (such as the piston, rings, cams, ...). We use these tools regularly in our consulting activity. These CAE programs are also widely used by the rest of the IC industry. WAVE is used at GM and DaimlerChrysler (two companies mentioned in the messages) as well as at most other OEMs and component suppliers worldwide. Interestingly we have also been approached by very prestigeous Research Labs in the US with the request of supplying our codes for research purposes (as part of our University Grant program).

Undoubtely there are advantages to linking 1D engine cycle simulation softwares (such as WAVE) to 3D CFD codes. WAVE has been linked to our own 3D CFD code VECTIS for the past 3 years and within a couple of months the interface between StarCD and WAVE will also be released.

If you need more information regarding Ricardo or WAVE please visit our website at www.ricardo.com or feel free to contact me directly.

Best Regards

Dott. Enrico Bradamante

voice: +1 (630) 789 0003 x 214 e-mail: EBradamante@ricardo-us.com

I don't exactly buy what you are trying to say. Recardo have hands on experience in designing and testing because the company has been working with the IC engine over and over again similar to Proving Ground type work of an automotive company. You are saying that Recardo is a consulting firm working for decades developing engines. Then tell me why are the engine companies getting so much penalty for producing engines with emission problems? What does it prove! Recardo or other consulting companies don't have the advanced technology to confront future stringent emissions requirements. Why can't you admit the truth. Truth is what we see with our own eyes. If the emissions control would be so easy problems then the President and Defense Secretary would not allow National Laboratories and Leading Universities facilities to utilize super-computers to perform computer simulations to tackle this kind of problems. 'WAVE' or other commercial softwares are mole-hills in a mountain and have limited applications. I am repeating again that this is not a child's play anymore. It may be in the past but not anymore.

Hi Mohammed,

I am interested in more details on these legal implications nad the example that you cite with the auto companies. I am in Canada and we are less apt to lawsuits but it is clear that there will be liability implications in the future. It is my understanding from several of the code delvelopers that there is lots of "small print" in the code liscence that basically says the code is not suited to do anything and the code developer is not liable for an idiot who uses it without adequate training in both the code mechanics and the general area of fluid mechanics and transport, etc.

Do you have any references (newspaper articles, law articles, etc.) that deal with the cases that you mention!

Thanks for the interesting and timely topic!

Duane

So what exactly are your National Labs and Superuniversities working on and what are they running on their Supercomputers?

Are they able to do a DNS for a complete engine with all surrounding parts including combustion?

Over the last 10 years there were several european programms funded by the EU and some automotive companies to find out a little bit more about the physics in engines.

Do you really think that a "Leading University" just has to switch on a supercomputer and suddenly becomes lightyears better than all others?