Centrifugal pump base fabrication

[JohnWeal]

Good morning,
We are in the process of installing a new centrifugal pump which will be fixed to an existing steel support structure.
I have read somewhere that the mass of a concrete support plinth for a centrifugal pump should be 3 x the pump mass. The pump in our case is 4600kg and the motor is 5875kg. (the motor is mounted on a support frame which also sits on the support structure)
Also I have read that two lines should be projected downwards at 30 degrees from the vertical and that the two lines should pass through the plinth base and NOT the sides.
In this case, we don't have a plinth but a steel structure.
Do the same rules apply?
Is there a different set of guidelines for steel structures as opposed to concrete plinths?
I am concerned about possible vibration issues which are specified at 11mm / second RMS.
The max speed of the pump is 750 RPM.
Please see attached sketch.

Regards
John

 

[bimr]

When a pump is mounted directly on a structural steel frame, tha pump should be located directly over, or as near as possible to the main building members, beams, or walls. The pump should be bolted to the support to avoid distortion, prevent vibration, and retain proper alignment. The installation should also be reviewed with the manufacturer.

If a Sub Base is being bolted to a structural steel foundation, use shims for leveling the plate.

Structural steel is much more flexible than a concrete base. Consideration also has to be given to the natural frequncy of the support structures.

 

[JohnWeal]

Hello mimr,
Thanks for the reply.
When you say that 'structural steel is more flexible than concrete' I assume this is a less satisfactory condition and therefore the steel frame needs to be made as rigid as possible.

This is my concern in terms of the pump having a high C og G on the frame.

Regards

John

 

[BigInch]

When mounting on floors, especially floors composed of steel beams, not as much "stiffness" is available as it would be if mounted on a typical concrete foundation & pedestal. The lighter steel sections do not provide much moment of inertia, thus deflections are generally much greater, and vibrations can be severe. Especially true when the natural frequency of the beam or floor is within 0.7 to 1.4 times that of the pump, in which case natural resonance amplifying small vibrations can reach relatively great amplitudes which can shut down the pump, if i+t has vibration sensors, or can cause extreme bearing stress and fatigue.

"People will work for you with blood and sweat and tears if they work for what they believe in......" - Simon Sinek

 

[desertfox]

Hi John

The purpose of the concrete plinth is to absorb vibration when the pumps running, I presume the manufacturer as provided you with an installation manual with reccomendations of how it should be installed.

See this link for a typical installation-

http://www.sureseal.co.za/pumps/Pump_Installation_Guide_LLC[1].pdf

and another link for design info

http://cbs.grundfos.com/export/site...iles/L-CBS-WP-02_Baseplates_FINAL_Low_Res.pdf

 

[desertfox]

Hi again

Re-reading you mention vibration velocity of 11mm/s, I've just come off a project were I oversaw about 20 pumps being installed, so driven with electric motors and some with diesel engines and we kept the vibration velocity to 1-2mm/s for the electric driven and a maximum of 3mm/s for diesel driven.
If you haven't spoken to the pump manufacturer then I would do so.
If you can tell me the power rating of the pump/drivers I have a chart here that installers work too for vibration velocity for a given size of machine.

desertfox

 

[JohnWeal]

Hello Desertfox,

The motor size is 500kW....

Regards

John

 

[desertfox]

Hi John

I'm in a rush but a quick look at the chart the max should be about 3.5mm/s.

 

[JohnWeal]

Thanks Desertfox...

Did you actaully view the drawing i attached in my original post? I have worked out the mass of the pump stool and it is 5400kg. The feet are bolted down to a 1m think reinforced slab.
Our new pumps are 6000 kg plus 6000kg of motor and integral motor support frame bolted to the stool. Hence 12000kg sitting on a 5400kg steel support structure.

I would prefer to remove the stool and pit some quality concrete in its place...

 

[desertfox]

hi John

I did have a look at the picture you posted, I would suggest that contacting the pump manufacturer and studying the installation manual provided might be a good starting point.
Is the new pump different to the old one? I notice that its an existing mounting frame and it might be that there are no vibration issues.

 

[GaTechTheron]

Some items to consider when designing the pump baseplate frame / foundation. Note that this discussion is assuming that we are dealing with a horizontal pump with bearings locating that shaft within the pump case. (Motor is not taking pump thrust.)

Pump Foot Forces on Baseplate:

1) Inertial Force as a Result of different Size Nozzles on Suction / Discharge: This can usually be neglected unless you have a very high flow rate pump.

2) Nozzle loads that you intend to put on the pump.

3) Reaction from drag within the pump: One pump foot/feet will have a fraction of pump weight PLUS moment reaction from parasitic drag on internal pump case parts, the other foot/feet will have a fraction of pump weight MINUS moment force from parasitic drag within the pump. The items that add to drag within the pump are as follows:

A-Fluid Friction Loss
B-Bearing Friction Loss
C-Drag From Seal Faces, Pumping Rings, etc

4) Force from thermal growth

Note that the torque applied to the pump shaft from the motor is NOT responsible of any pump foot reaction. HOWEVER, it may be good practice to design for the "locked rotor" case. In other words, if the pump is locked up, the motor will transmit all of it's locked rotor torque to the pump, and the pump will (in this case) transmit this torque directly to the pump feet.

5) The motor will have an equal torque reaction on the baseplate to how much torque it is delivering. This adds to forces on one side of the motor, and subtracts from the other side.

Pump Vibration:

From a macro standpoint, it is a good idea to consider the TRIP vibration level. Say you plan on shutting down the pump at 0.30[in/sec]. If you know the MAX frequency of this vibration, you should be able to calculate the acceleration of the rotor under max vibration / max frequency scenario. Using acceleration and weight of the rotor, you can calculate the force that will be transmitted to the pumps bearings and therefore the foundation. Calculate this force assuming 100% tranmissibility, add this to the weight of the pump when full of liquid, as well as nozzle loads, and applicable inertial forces (item #1 above) and other items above, and you have a rather strong idea of the maximum load that the foundation will see.

This method, however, can be somewhat laborious.