Foundation Natural Frequency 1

Depends on how sensitive the piece of equipment is, really. Typically for a typical industrial motor/pump application, I will design the foundation 3x the weight of the motor.

Alternatively, if you are designing a foundation for a generator (high $$ equipment) you may want to consider the soil properties as well. A good reference book would be Design of Structures and Foundations for Vibrating Machinesby Arya, Oneil and Pincus.

The book AUCE98 suggests, by Arya, O'Neill and Pincus, is excellent.[ ] As well as a lot of useful general information, it gives you a step-by-step approach to your particular problem.

SlideRuleEra has a document on the subject in his library

Thanks for the quick responses. It so happens that I found a document on civildesignhelp.info that refers to the book by Arya, O'Neill and Pincus. It claims that if the motor is less than 200HP, (mine is 75), and weight of machine, (assuming "machine" in this case is pump plus motor), is less than 5620 lb, (mine is 4175), then dynamic analysis is not required. So I think the 3 times the weight of machine in concrete is probably the rule of thumb. (Our preliminary foundation design weighs about 35 kips). So I think it's going to be fine. However, my curiosity is piqued - how to determine the natural frequency of a block of concrete?? It sounds absurd! There must be some wacky mathematical formula for that, eh? Anybody know?

The "wacky formula" (formulae actually), based on the theory of an elastic halfspace, is/are given in AO&P's book.

An intriguing question, and I'm sure the reference stated is the best bet.

It occurs to me that you might have a reasonable estimate of the spring constant of the foundation, but finding the effective mass (concrete plus effective weight of soil that moves with it) would be the problem. If you neglect the soil weight and figure the natural frequency without it, that should overestimate the natural frequency. If reducing it would move it farther away from the frequency of your motor, that's all you really need to know. Or, if it is a lot different, that's probably close enough. But if the frequencies are in the same ballpark, or adding additional mass would move them closer together, then it would need closer analysis.

It also occurs to me that you could have natural frequencies vertically (which I assume is the main issue), laterally, and rotationally in three directions. There, the book would be useful to see what modes ought to be checked.

penpe - See the following documents:

"Don't Gamble on Machinery Foundations"
"Principles of Foundation Design for Engines & Compressors"
"Basic Vibration & Vibration Isolation Theory"

on this page of my website:

it seems you are new to petrochem industry. you probably work in a small company because all big petrochem engineering firms has design guides.

fluor has a good design guide for pump foundation. the formula for natural frequency for pump foundation is


Fn = 60/2phi * sqrt (k/m)

It is probably a bit late for the OP by now, but this thread inspired me to tart up one of my old spreadsheets, one that implements and adds to the method given in the Arya, O'Neill & Pincus book.[ ] The result is available for download from my web site (