Transient vibration of an engine

[vco]

How is the transient vibration of a piston engine usually simulated?

I know that in order to define the vibration loading you need the mass properties and dimensions of the moving components plus the cylinder pressure curve. And of course you need to know the firing order, V-angle (if applicable), etc.

However, since this is a transient case you would in fact need a set of cylinder pressure curves for different engine speeds and throttle positions, right? Depending on the chosen resolution for the engine speed and throttle position that sounds as quite a lot of data to be measured. Are there any industry practices where you utilize clever interpolation and thus avoid having to obtain such a vast number of cylinder pressure curves?

I am not especially experienced regarding engines, but I am very familiar with vibration analysis in general.

 

[anorlunda]

How is the transient vibration of a piston engine usually simulated?

How do you define transient vibration? Do you mean the instantaneous position?

 

[vco]

How do you define transient vibration? Do you mean the instantaneous position?

By transient vibration I mean that the engine speed and torque are changing as a function of time, as opposed to steady-state vibration where the engine is being run at a constant speed and torque.

Well, to be exact the instantaneous values for the engine speed and torque are changing in both cases, but I guess you know what I mean.

 

[Dr.D]

The simple answer is that this is rarely done and has no well established procedure. As long as nothing breaks in the transient condition, the engine state will soon be different and it did no matter anyway. Even if you know every physical detail about the engine (weights, dimensions, pressure curves, etc), you still do not have enough information until you know the load in comparable detail.

Engineers tend to worry much more about steady state vibrations. This is the motion that causes fatigue damage and also noise. It goes on and on (that's what steady state is all about!), and that is what is damaging and annoying.

 

[vco]

The simple answer is that this is rarely done and has no well established procedure. As long as nothing breaks in the transient condition, the engine state will soon be different and it did no matter anyway. Even if you know every physical detail about the engine (weights, dimensions, pressure curves, etc), you still do not have enough information until you know the load in comparable detail.

Engineers tend to worry much more about steady state vibrations. This is the motion that causes fatigue damage and also noise. It goes on and on (that's what steady state is all about!), and that is what is damaging and annoying.

I would imagine the procedure is indeed relatively rare.

But what about motorsports, for example Formula One? Surely the teams perform such analyses since the engines are rarely, if ever, running in steady-state conditions and fierce competition requires simulations to be as realistic as possible?

 

[Dr.D]

Not being a motorsports fan, I really cannot address the question from experience. But I would suggest that this is one of the least likely places to find such analysis. The details of such an analysis are difficult, tedious, and take a lot of time. Most motorsports are done with "seat of the pants" engineering, not with extended expensive computer analysis. In that arena, there is much more of a "cut and try" approach. If it does not blow up, it must be a success. So again, I doubt that this is done with any frequency at all.

 

[cjl]

I would imagine the procedure is indeed relatively rare.

But what about motorsports, for example Formula One? Surely the teams perform such analyses since the engines are rarely, if ever, running in steady-state conditions and fierce competition requires simulations to be as realistic as possible?

It's probably closer to steady state than you think. Keep in mind that the engine is completing a full revolution hundreds of times per second, so even very fast throttle transients are pretty slow relative to the engine's performance. I'd guess that the most important transient that they simulate is when the engine changes RPM dramatically on a timescale of tens of milliseconds every time the car shifts, but even then, I'd imagine it's more an issue of shock load and momentary extreme load rather than vibration.