DynamicTime Analsis and UNITS?

[cbra]

I have performed a Dynamic Time Analysis which has (4) Acceleration Measures. My input is G^2/Hz vs Frequency. When I plot the results from the acceleration measures vs frequency I am not sure what the units are (What vs Frequency).



Are they in the same units as my input G^2/Hz or just G's?



Also, is there a way to import the graph reports into a microsoft office product?



Thanks in advance for you help!



Brian

 

[JHardy]

I am not familiar with your unit of acceleration (G^2/Hz), so I am not surprised you are having trouble with your units!



Dynamic analysis requires you to use consistent units throughout - once you have chosen your basic unit of mass (eg kg), length (eg metre) and time (eg second) all other units are effectively specified.



Unit of acceleration should be (Length/Time^2) - eg m/s/s or ft/s/s or inch/s/s. The g is a commonly used unit of acceleration (approximately 9.81 m/s/s), but should not generally be used in FEA, because it is not consistent. It needs to be scaled into consistent units (such as m/s/s).



Yoir unit of G^2/Hz would have units of (m^2/s^4/s^-1) = m^2/s^3, which has no practical meaning that I can think of.



Hope this helps.

 

[JHardy]

Brian,



I have re-read your original posting, and I realise I missed something the first time around.



I assume you are carrying out a spectral response analysis for random excitation. It sounds like you have been given a “Power Spectral Density” (PSD) curve. PSD analysis is commonly used to determine the response of a structure subjected to a statistically averaged random excitation. The random excitation can be a force, acceleration, velocity or displacement.



In the case of PSD the loads applied to the model are expressed as a power spectral density. Typical units for the different types of excitation spectra are m^2/Hz for displacement, (m/s)^2/Hz for velocity, (m/s/s)^2/Hz for acceleration and N^2/Hz for load spectrum. The acceleration spectrum is commonly expressed as g^2/Hz (g=acceleration due to gravity). In this case the actual magnitude of g^2 (e.g. 96.2361 m^2/s^2) must be included in the direction vector.



I don’t honestly know what units your output will be if you forget to include the scaling factor to get back to the basic units of (m/s/s)^2/Hz. I suggest you include the scaling factor in your analysis, and then you can be sure of your units.



I have to admit that I don’t perform this type of dynamic analysis as a rule, so that just about exhausts the extent of my knowledge on PSDs! (I don’t even know whether Pro/Mechanica even supports this type of analysis!)



Hope this helps!



(Much of the above advice came from the Help file associated with a fine Australian FEA package I use - Strand 7 – website www.strand.aust.com )

 

[ray]

Brian



Following on from the other replies, a Dynamic Time analysis will expect the input to be units of acceleration, e.g. m/s2 or mm/s2. If you have an input in G2/Hz, then that is Acceleration Power Spectral Density, which is a unit used in Dynamic Random Analysis. Pro/Mechanica can do Dynamic Random Analysis, as well as Dynamic Time, Frequency and Shock Analysis.



Therefore I think that you may need to study exactly which kind of analysis you really need to be doing in your situation. The different types of vibration analysis, when to use them and how best to set them up and interpret the results can be confusing, and probably requires a longer explanation than can be sent as a reply in this forum. This is why at Elite Consulting in the UK we run a 2 day advanced training course on Vibration.



You can look at our web site for more info www.elite-consulting.com or go to the Customer Support area of the PTC web site and search the knowledge base.



Regards



Ray Ellender

 

[cbra]

Thanks for the responses! They helped me out greatly. I seem to always get messed up with units in mechanica.



I wish they would incorporate assistance within the software when setting up the analysis.



Again, Thanks!



Brian

 

[devarajan]

hey every1,



i have quite a lot of doubts in vibrational analysis..... it might seem trivial to you, never'less



1. brian --- how do you create acceleration measures?? and what is it used for ???



2. when i apply a base excitation in dynamic time analysis .. what units is it in ?? Hz???? newton ????



3. i have a spring in my analysis . i want to perform a sensitivity analysis ( by varying the spring constant ) .. but am not able to create a Design parameter for the spring constant ....



can you suggest me a good tutorial for vibrational analysis in mechanica .........



thnx in advance



deva