How to determine exhaust gas induced vibrations? 3

[jean15paul]

Hi all,

This is my first post in this forum.

I have a problem that I'm not sure how to approach. I'm going to be doing an structural and thermal analysis of an engine exhaust system. The scope of work specifies that I need to consider "exhaust gas induced vibrations." Does anyone have any insight as to how to determine a vibration load caused by the exhaust gas? Is it just the exhaust gas pressure, cycling at the engine firing frequency?

(Background on me, feel free to ignore) I have ~5 years of experience in stress analysis of primary structures (classical and FEA), but everything I did at that job was linear static. We had a seperate department for dynamics/vibrations. I just started a new job where I'm responsible for both static and dynamic FEA analysis of structures, so I'm pretty overwhelmed trying to learn all the dynamic and vibration stuff. So I'll probably be posting questions on here often.

Thanks in advance for the help.

 

[ajamnia]

I would say gas pressure cycling is a factor as well as thermal cycling that you would get. Depending on your geometry, you may also have induced vibration due to aerodynamic effects. Are there some old timers that you can discuss the physics of the problem with first?

 

[jean15paul]

Thanks for the help ajamnia. That makes sense. Unfortunately I don't have the aero-fluids background to figure out the aerodynamic effects.

One thing I'm struggling with at this new job is that they don't have any old timers that I can ask questions. My last job was full of them, and they were always a great resource. Now I'm at a much smaller place. I have much more responsibility and much fewer people that I can guide me, so I have to figure out a lot on my own. That's the main reason that I joined eng-tips.com.

 

[BrianE22]

I have no experience with this but I think I'd start by installing pressure transducers in the exhaust at various positions. Also possibly accelerometers on the piping and maybe a microphone. Collect data at different engine speeds and temperatures and look for correlations in the data.

 

[GregLocock]

" Is it just the exhaust gas pressure, cycling at the engine firing frequency?"

Sort of. The pulses contain many different orders (multiples of crank speed), starting typically at half order.

There is also gas rush noise which occurs when the flow goes turbulent, and some other effects which I'll leave out for the time being.

The main one that people forget about is that the engine itself is moving and its vibration is transmitted to the exhaust system.

Needless to say the frequency content and amplitude of the pressure vary along the pipe, fundamentally it is an organ pipe.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[DLite30]

Is this for a turbine or a reciprocating engine?

Either way, I've never heard of this done for any kind of machine, except for silencer manufactures who may do this kind of study in some cases to predict shell vibrations.

Typically the exhaust piping is somewhat isolated from mechanically induced vibration from the engine by some kind of expansion joint or bellows, so that's not generally an issue.

I would consider just looking at any existing natural frequencies of your exhaust layout, and determining whether they closely coincide (say within 10%) with any exhaust pulse frequency from the engine (for a recip engine). I would imagine that the engine vendor could help you with determining those frequencies.

For a turbine it's essentially a non-issue, as it's a continuous, non-pulsing flow. Maybe consider some kind of vortex shedding issues, but you shouldn't have anything significant engough in the exhaust flow path to cause that. But you may want to go ahead and determine the MNF's for the exhaust system, so you have some sort of work product.

I would also consider external factors such as vortex shedding of ambient air around the silencer at design wind speed conditions.

My 2 centavos.

 

[SomptingGuy]

Sounds like coupled fluid structure interaction to predict shell noise to me. Exhaust systems don't fail because of gas loads. Thermal should be fairly simple as the frequency of the gas pulsations will be well above the thermal response time, such that cyclic mean temperature can be used.

I assume this is a transient situation too, so that the excitation frequencies will vary with engine speed.

- Steve

 

[MikeHalloran]

As Greg says, recips wobble on their mounts, particularly during startup and shutdown. Proper exhaust systems include at least one bellows, and the system itself is often suspended on springs, to limit structureborne vibration and also the amount of heat conducted to the structure.



Mike Halloran
Pembroke Pines, FL, USA

 

[FeX32]

I agree that it is a FSI (fluid-structure-interaction) problem. You may consider a multiphysics FEA package.


Fe (IronX32)

 

[Tmoose]

"Exhaust systems don't fail because of gas loads."

2 stroke engines with flat sided expansion chambers are generally much noisier than those with round chambers due to chamber walls flexing in response to the pressure waves. The flat chambers that Yamaha provided to fit on the original TZ-750s failed by cracking and were the source of some DNFs.

Simlarly, HVAC duct designers recognize the big "break out" advantage round ducts have over rectangular ducts even with the relatively tame pressure variations they deal with.

 

[jean15paul]

Wow. Thanks for all the insightful responses everyone. Being new to eng-tips.com, I'm very impressed how helpful everyone is.

This is an propulsion system for a new ship. It hasn't been built yet. I was recently asked for a time estimate to do this analysis whenever the project get started. I don't have the information necessary to get much deeper into this problem, but let's see if I can respond to some of the questions/suggestions that were brought up.

Since it hasn't been built yet, I can't get test data. (Test data is always best.) Although we've built a similar ship in the past, so I'll try to see what was done for that one.

I don't know if the systems will be mounted on springs/isolation (if not, I may suggest that), but I do know that an expansion joint is planned.

I'm not sure if it's a recip/turbine engine or a piston engine. But it definitely is a transient problem as the engine won't always be operating at the same speed. Good insight on the thermal response time.

Structures-fluid interaction, vortex shedding ... I don't have the fluids backgroud to even begin to know how to address these.

Thanks again.

 

[JJPellin]

I believe that the last large diesel engine project in our plant include a requirement to analyze the exhaust piping for the possibility of standing wave acoustic resonance. I will try to find the results of that analysis. If I can find that report, I will summarize the sort of analysis that was performed.

Johnny Pellin