How to calculate dynamic loads for a PD blower 1

[hirschaplin]

Hello,

When designing PD blower packages there is always a returning document deliverable that disturb me a little bit:
"Civil and Foundation Drawing with total weight, anchor bolt location dynamic forces and moments"

The disturbing part is the dynamic forces and moments as it seems like none of the big PD blower manufacturers provide clear guidance on how such loads shall be calculated.

Then I came across a foundation drawing that was made by Howden Roots for a single package with two blowers driven by a common electric motor - and to my surprise they provided a table with static and dynamic loads!

Overview of the skid with load points indicated:

Load table:

Weight table:

Then I was recently involved in a project with a big European contractor and they have a specification for what input data they require to be supplied in order for them to design a foundation themselves for rotating machines, with the following scope text: "THIS STANDARD RESUME THE FLOW OF INPUT DATA NEEDED TO CORRECTLY DESIGN THE FOUNDATION OF ROTATING MACHINERIES".

Very interesting, I was expecting to finally get some clarity on how to handle the issue with dynamic loads - a general approach that I can apply in all my projects if you want. But no, the darn specification only show which data they require, not how to use it:

No clear category match for PD blowers, but I would assume it would fit best under "7. PUMPS - TURBINES - FANS AND TURBOCOMPRESSORS.", do you agree? I guess with all this input data, they have enough to calculate all loads by themselves. How are they doing that, how can I use this input data to establish relevant dynamics loads for any PD blower I might use in my projects?

In my packages, the blower is most often direct driven but occasionally we use v-belts. Will the dynamic loads be different for the same machine and operating data but with different driving method?

I am not super concerned by my lacking skills... Today we design our foundations without calculating dynamic loads, just going big and bulky. It is working but I don't like to just follow my guts feeling, I want a decision to be driven by facts or engineering data because I guess it is just a matter of time when something actually goes really wrong with this current approach. But I am also not sure how big of a problem it really is as it seems like dynamic loads are pretty low for PD blowers.

I recently did a FAT of a 250 kW blower. The skid was standing on top of 6 small 5 mm dia. spot supports, 35 cm above floor level, without being locked in place with bolts or something similar. The skid didn't move a millimeter at full load and vibration levels at around 3-4 mm/s RMS on blower casing above bearings. No vibrations in the skid structure.

Loads on the Howden Roots drawing are quite big numbers, but it seems like maybe the loads are taking out each other as they are same on both sides and therefore not very impactful to the design of a concrete foundation? Please elaborate...

But how to approach it? It is awkward to NOT provide dynamic loads or have a good way to talk about it.

 

[1503-44]

Dynamics are usually going to be minimal for 250-500Hp range rotating machines.
Reciprocating pumps and compressors generate more unbalanced forces, thus more problematic.
That may be why you are not getting any data. The mfgr didn't calculate it.
You can estimate forces from weight and power and rpm data of the machines.

In the lower power ranges less than 500 Hp, foundations are usually designed to minimise weight eccentricities on the foundations, which are then sized to provide enough mass to damp out any vibrations simply by providing 3 to 5 times the mass of the machines.

If you get into larger power ranges, the dynamics become more important and full dynamic analysis is done. You get the spring constants of the soil below from your geotechnical shear wave studies.
Weight loads and operating torques of the machines are then required to be furnished by the manufacturers for final foundation design, but can be estimated for initial design using machine dynamic principles.
Torque = Hp * 5252/ rpm
Anchor bolt loads for Torque are torque/number and distance apart.
Total Bolt load = P/N (static) +/- torque loads (your "dynamic" forces above).
You need to obtain rotational moments of inertia of machine and driver for calculating runup dynamics and locked rotor torque conditions.

Foundation response and vibrations are done by calculating the natural frequency of the fdtn and keeping operating frequencies outside 0.75 to 1.4 Fn. Keep vibrations small and operating times short within that range.

Once you get your soil spring constants, it's all just Dynamics 101.
Rotating machines are relatively easy. Recip equipment is more of a PIA.

https://www.researchgate.net/publication/306498089_Dynamic_Analysis_Of_Machine_Foundation

Example Calc ...

https://vdocuments.net/block-foundation-dynamic-analysis.html?page=13

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[hirschaplin]

Thanks for your reply but I feel it is on a very high generic level. I want to go on the depth for PD blowers specifically to get something out from this thread that I can use to improve my foundation drawings. How is it handled, how did Howden Roots get to their static and dynamic loads that I am showing above?

Why is the european contractor I mentioned asking for those parameters specifically and not other parameters? For example, they are not asking for the moment of inertia? How is their calculation using those input parameters?

I don't think it is as easy as how you suggest and the reason is that there are an infinite amount of design options. For example, if you have a skid with 4 HEB legs supporting a machine table where the blower and motor is installed I guess some of the torque will transmit down into the concrete foundation under the skid. If you add 4 anti-vibration mountings (AVM) between the HEB legs and machine table, will the loads into the concrete be the same, reduced or eliminated:

Your estimation idea is also not considering nozzle loads. If I have a 500 hp blower there will defiantly be some big pipes coming into the skid that might provide loads that must be handled by the concrete foundation.

Designing the foundation to have 3 to 5 times the mass of the machines is also a classic statement that I have heard many times before and even used on my own drawings. Last time I used it my customer asked me if they should make the concrete foundation 3 or 5 times the weight of the machines. He asked me to clarify why we provide a range and how he could know whether to choose 3 or 5 times the weight. I couldn't explain the range, like why he have the option to choose form 3 to 5. It is strange and I felt a bit silly for putting my signature on something I didn't fully understood or could explain.

The 3 to 5 mass of rotating machines is also a bit too simple. If the design is using a similar design as here with AVM's... is the machine table (highlighted with red below) itself suppose to be 3-5 times the weight of motor and blower? If yes it is almost impossible even if you fill the table with concrete, I think it will only work if you really oversize it.

Or are we talking about the concrete foundation hosting the full package?

I would guess all dead weight of the package should be reduced from the 3 to 5, meaning that if the complete package is 10.000 kg - 2.000 kg of that is rotating machines then 3 times would be 6.000 kg, but the overall package already have a surplus of 8.000 kg dead weight over the rotating machines to damp out any vibrations. In that case I guess the concrete foundation does not need to be 3 to 5 times the total weight of the package (30.000 to 50.000 kg)?

 

[1503-44]

The "mass rule" is based on rated power and type of machine and is limited to small power capacities. 250 to 500 Hp max. Anything more, and you do a dynamic analysis. It's just based on experience. Rotating equipment 250Hp = 3xM, recip 500Hp =5xM The resulting mass is the weight of the foundation, pedestal blocks, footing block and would also include soil, if there is any soil above a spread footing extending past the pedestal blocks.

The loads in the table above are probably derived from max torque of each machine component and the weight of each component above/4 AB.
Estimate load from 2057 component
Assume machine weight of 7032 Kg divided by 4 Anchor Bolts, B1, B2, C1, C2
plus the base frame load /14 AB
7032/4 + 4550/14 = 2081 Kg Well that's pretty close to the posted bolt line B static load and only 31Kg different from the C bolt line.

To estimate the dynamic forces at the base, one would have to know the torque being applied, or power and rpm, and the height above the base of the rotational axis. Then those could be translated to forces/moments about the base and then to anchor bolt loads.

Rotational moments of inertia are needed to calculate forces during rotational acceleration and time to reach operating speed, if that is important.

If springs are used for vibration isolation, they have no effect on static loads.
They deflect by an amount that corresponds to the torque being produced according to their spring constant, transmitting the resulting spring force to the skid/foundation. They really do not have any effect on changing the torque load transmitted to the skid/foundation when torque is constant.
While rotation is accelerating, they allow some rotational displacement to occur which is counteracted by the machine's inertia, rather than being transmitted to the skid/foundation. Since acceleration only happens during start and stopping of the machine, they are only active at that time and during accelerations caused by unbalanced rotor forces, which can cause vibration.






--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."