Frequency-domain analysis of valvetrain dynamics

[ivymike]

Hello

Anyone ever muck about with frequency-domain analysis of a valvetrain? I'm currently working on some analytical tools to help me with said analyses, and I'm wondering if there might be some pointers that someone could throw out for me, as I'm a bit green in this area. Most of my experience w/valvetrains is via time-domain simulation.

If you've looked at valvetrain vibration in the frequency domain, I'd be interested to know more about what you've done, and what parameters, etc., you typically look at. A brief description of your methods would be most appreciated.

Thanks
Isaac

 

[buzz41]

Years ago, when I looked at the cam ring of a large radial aircraft engine, a little light came on in my noggin, I recognized the math behind what was used to generate that profile. It was plainly a sin^2 pulse with the 3rd harmonic only, maybe similar to sin^2 + 1/3*sin3^2. It even had a little dip in the middle. I thought, so thats how they do it to minimize the wear and tear on the valve train. That cam was probably designed in the 40's. Seems the valve boys and their frequencies have been at it a long time.

Isaac, I tried my hand at doing a vibration analysis of a crankshaft using EE tools some time ago. I was a bit clumsy at it, but got what I wanted. SPICE is one tool that is available as a freebee in lite versions. Various functions are there - Fourier to do freq/displacement conversions, etc.

I think SAE has some nice dedicated software tools but they don't seem to know the word free. $$$$ :-(

 

[ivymike]

Well I've got some pretty sophisticated tools at my fingertips as it is - the key is getting the precise information that I want to look at (and then teaching the software to give me only that info, instead of volumes of useless stuff), and at this point I'm not entirely convinced that I know what that is. I do know some fundamentals, such as mode shapes and resonant frequencies, etc., but the part I'd like to know is what to look for in particular, and what techniques can be used to solve vibration problems based on frequency-domain analyses.

It seems to me that for some valvetrain configurations, an improvement in vibratory behavior might be achieved by redistributing inertia & stiffnesses (when constraints prevent broad changes to hardware). A perfect example - is it better to have a heavy, stiff pushrod, or a light, flexible one? The valve spring can only control so much mass, but too much flexibility is bad too (and you don't want to buckle the thing, etc). I was hoping to be able to find answers to questions like "what is the right mass&stiffness for a pushrod in this system" by looking at the mode shapes & natural frequencies. If anyone has some general pointers in that direction, I'd appreciate them.

Also, it seems to me that the appropriate way to represent the system excitation is via a fourier series expansion of the cam acceleration profile (since excitation force will be roughly proportional to the acceleration). I'd assumed that the right thing to do was "expand" the whole cycle (including the base circle), but someone suggested that I might want to look only at the portion between the start of the opening ramp and the end of the closing ramp. If someone had previously performed the analysis, perhaps he wouldn't mind sharing how he represented the excitation?