Skip to main content

Replies sorted oldest to newest

Tony,

Some of your ODS works are impressive. I have a few questions for you: Do you make all your models in MEscope or do you use other 3D modeling packages ? Do you make measurements at every joint or just a few and rest are interpolated ? My models, due time constraints are most of the time stick (line) models. In your "Plunger pump and piping" example, roughly how many triaxial measurements do you have? Thanks
RM
Hello Joe. Thanks for the quesitons.

I make most of my models in ME'scope. Everything you saw on my website services page for ODS and Modal was done in ME'scope. On occasion, I have taken an autocad 2D or 3D drawing and imported into ME'scope and modified the object to fit my needs.

I try to keep the drawings as simple as possible, yet I want the structure drawing to look like the machine. On a rectangular structure like an A-frame or concrete base, I make measurements at every joint (every place two lines intersect in the drawing). For a cylindrical structure like a motor or pipe, I take one X, Y and Z direction measurement at a cross section, asssign it to the radius point of the cross section and let the software interpolate the unmeasured points. The assumption I'm making here is that I'm only interested in lower frequencies where the pipe or motor housing remains rigid as it is vibrating. If I was measuring out to 2000 Hertz on a pipe or other cylindrical object, I would need to measure points around object because I know the pipe is probably deforming at the higher vibration frequencies.

In the plunger pump example, there were 450 points tested with three directions at each point for about 1400 total measurements. That's more points that I ususally need to test but I needed a lot of points because of the piping.
RM
Dear Tony

The ODS animations are really good. I am not an expert on ODS or Modal analysis. I always wonder that how, ODS, which I believe ( pl correct me if wrong ), only shows how a machine is moving ( dancing ) help a customer to solve structural problems. It does not give me nodes / antinodes which are required to decide where to stiffen and where not.

MEscope has some advanced softwares which also gives solution as to where and howmuch stiffening is required to solve problem.

What I want to know is that my merely animating the machine movement, can we effectively solve complex structural problems. Modal analysis/ODS services is, I believe, more expensive then any Pdm program charges and does require substantial amount of experience.

Regards
RM
Hello VRG

ODS testing is able to identify a wide variety of mechanical faults and resonance issues such as looseness, soft foot, broken welds, misalignment, unbalance, bending resonance modes, torsional resonance modes, structural weakness and foundation problems. The ODS doesn’t care if the machine problem is non-linear like a broken weld or loose bolt. ODS shows everything that is happening on the machine.

The ODS animation will show the antinode lines in the animation when you turn on the contour colors if that is helpful to you.

ODS animation will not always be able to help you correct complex structural issues. ODS animations will show you how the machine is vibrating, bending or twisting which may lead to ideas on how to modify a structure to move natural frequencies away from forced vibration frequencies. You could do a modal analysis, but like an ODS, the output of a Modal Analysis does not indicate how to modify a structure. It only shows you all of the natural frequencies, mode shapes and damping values that were excited in the test. One more thing, a Modal Analysis assumes that the structure is linear. What if it is not linear? The structure could have broken welds or loose components.

The best way to fix a resonance issue is to first do a modal analysis and then use the structural modification tools in the modal software (if you have them) to investigate various structural modifications.

Another best practice path that can be followed to correct resonance problems is to first do the modal analysis then contract a structural engineer to do a Finite Element Analysis (FEA) of the machine. The engineer can use the modal test results to fine tune the FEA model so it will produce accurate results when used to estimate the effect on natural frequencies when structural modifications are applied.

Having described the "best" way to pursue correction of resonance problems, we can now talk about other options…

After studying the ODS or Modal animations, it may be cheaper and faster to use your mechanical intuition and industrial experience to figure out how to change the mass or stiffness of the structure and move natural frequencies away from forced vibrations. There is no guarantee that the modification will solve the resonance problem. The outcome will be one of the following:

A. The modification worked and moved the natural frequency away from forced vibration thus reducing the effect of resonance.
B. The modification worked but moved the natural frequency close to a different forcing frequency resulting in a new resonance problem.
C. The modification worked but now the modification itself is resonating.
D. The modification did not work.

Attempting to correct resonance problems using industrial experience and mechanical intuition is not a bad approach if it is inexpensive, quick and safe. Trial and error is not the best approach for every machine or resonance problem. FEA's can be expensive. I think doing a FEA to correct a resonance problem without first doing a modal analysis is a waste of money.

If I had my choice, I would follow this diagnostic path for resonance problems.

1. Complete route based or stand alone Spectrum and Waveform analysis on the machine. If the source of the vibration problem is not obvious in the data, additional diagnostic testing is required.

2. Operational Deflection Shape testing (ODS) can be completed to show how the machine is deforming during normal operation. The ODS doesn’t care if the machine problem is non-linear like a broken weld or loose bolt. The ODS shows everything happening on the machine.

3. If structural bending or twisting was identified in the ODS test, a modal survey is needed to identify the mode shapes for all natural frequencies in the frequency range of interest. ODS and Modal testing do not directly identify how to correct resonance problems. Sometimes the structural modification solution is obvious through study of the ODS animations. The β€œtrial and error” method should be considered if it is quick, inexpensive and safe. When the solution is not obvious, additional theoretical and/or experimental solutions must be considered.

4. The outputs of a modal survey are the natural frequencies, mode shapes and estimates of structural mass, stiffness and damping. Higher end versions of modal software contain finite element tools that can be used to evaluate proposed structural modifications. My version of ME’scope Modal/ODS software has these structural modification tools. Working with an engineer at the site, I would use these tools evaluate structural modification options. Alternatively, a structural engineer can be contracted to complete a Finite Element Analysis (FEA) of the machine or structure. FEA is a theoretical structural engineering tool used to correct resonance problems. The structural engineer will benefit from having experimental modal data to use to fine-tune the FEA model before using it to evaluate structural modification options.

I hope this helps. Please write back if you have additional comments.
RM

Add Reply

×
×
×
×
Link copied to your clipboard.
×