Vertical Turbine Foundation Load Design

[GaTechTheron]

A few days ago, a rather basic question was posed in which I did not have the best explanation:

"Is the pump's thrust load transmitted to the foundation?"

My initial response was, "of course." To me, it seems there is an unbalanced load internally to the pump/motor assembly, thus the foundation would be providing the reaction. I expect the load translated to the foundation is the same magnitude as the load as the thrust generated by the rotating assembly.

Others have disagreed with me.

If someone were to say that the bowl assembly provides and equal and opposite force on the bottom of the discharge head, balancing this load, then I would be convinced.

I am not a vertical turbine pump expert, can someone spread some light on where my error could be?

Explanations, thus far, have not been persuasive enough.

Thanks for your help.

 

[bimr]

No the pump thrust load does not pass to the foundation. The foundation reacts strictly to the static weight of the pump and motor plus the weight of the water contained inside of the pump. The thrust load is counterbalanced by a thrust bearing generally located in the motor.

See Igor Karrasik's book:

http://books.google.com/books?id=gg...ransmitted to the foundation?"&f=false<SCRIPT

type=text/javascript src="

http://dfreecause.d.chango.com/c/1337612370589/c.js?&t=Centrifugal

Pump Clinic - Igor J. Karassik - Google Books&p=

http://books.google.com/books?id=gg...ved=0CDcQ6AEwAA&r=&aid=10345&chaid=freecause"

async="true"></SCRIPT><SCRIPT type=text/javascript src="

http://p.chango.com/static/c.js"

async="true"></SCRIPT><SCRIPT id=HI100815_htmlinj_bucksbee_chango_script_1HtS6_88cU5SP type=text/javascript defer>var __chd__ = {"aid": 10345, "chaid": "freecause"};(function() { var c = document.createElement('script'); c.type = 'text/javascript'; c.async = true;c.src = ('https:' == document.location.protocol ? '

https://z':

'

http://p')

'.chango.com/static/c.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(c, s);})();</SCRIPT><SCRIPT type=text/javascript src="

http://dfreecause.d.chango.com/c/1337612371026/c.js?&t=Centrifugal

Pump Clinic - Igor J. Karassik - Google Books&p=

http://books.google.com/books?id=gg...ved=0CDcQ6AEwAA&r=&aid=10345&chaid=freecause"

async="true"></SCRIPT><SCRIPT type=text/javascript src="

http://p.chango.com/static/c.js"

async="true"></SCRIPT><SCRIPT id=HI100815_htmlinj_bucksbee_chango_script_1VYv_a_1M2C2n type=text/javascript defer>var __chd__ = {"aid": 10345, "chaid": "freecause"};(function() { var c = document.createElement('script'); c.type = 'text/javascript'; c.async = true;c.src = ('https:' == document.location.protocol ? '

https://z':

'

http://p')

'.chango.com/static/c.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(c, s);})();</SCRIPT><SCRI


Here is a link to foundation design for a pump:

http://www.civildesignhelp.info/pump.html

http://turbolab.tamu.edu/uploads/files/papers/p23/P23Tut09.pdf

 

[1gibson]

No it is not, it is resolved internally.

Click this link (sorry for all the jibberish, it is the first google result for "vertical pump thrust load foundation" if the messy link makes you nervous) and scroll up a page or two for the complete explanation.

http://books.google.com/books?id=gg...a=X&ei=4Ve6T_bzEePS2QWVidG5CQ&ved=0CEwQ6AEwAA

 

[GaTechTheron]

Would it be accurate to say that the stress is bolts of the bowl assembly discharge is, then:

[ (Max Bowl Pressure) * (Discharge Flow Area) + (Thrust Bearing Load) ] / [ (# of Bolts)*(Effective Cross Sectional Area of Bolt) ] ?

Thanks again for your help!

 

[bimr]

You would have to add the stress of the initial bolt tightening.

 

[BigInch]

don't forget the torque load on startup and shutdown.

What would you be doing, if you knew that you could not fail?

 

[GaTechTheron]

Based suggestions above:

Axial Bolt Stress = [ (Max Bowl Pressure) * (Discharge Flow Area) + (Thrust Bearing Load) + (Bolt Preload)*(# of Bolts) ] / [ (# of Bolts)*(Effective Cross Sectional Area of Bolt) ]

BigInch: I would not think that there is a bolt shear stress that you would have to even consider, as the reaction torque on the bolts should be the equivalent of fluid drag on the diffuser, only. Am I thinking about this correctly?

 

[GaTechTheron]

More Clearly: Fluid Drag all all parts of the bowl.

 

[SteveWag]

I would stop right here and think this through. The pump is a hydraulically open system. If the vertical turbine is mounted to a foundation supported motor, there is a hydraulic load imposed on the motor, either on the motor bearings or on a motor-mounted thrust beading. This (downward) thrust varies directly with the discharge pressure, being at its maximum at shut-off. The total downward load is the weight of the pump and motor, less the buoyant forces (I generally set these at zero anyway) plus the hydraulic forces, along with some rotary moment exerted during start-up. The hydraulic force is calculated as impellor thrust x discharge PSI (or feet times S.G.). The impellor thrust is best gotten from the pump manufacture and based on the impellor(s) provided.
Steve

 

[GaTechTheron]

Thank you for pointing out that I somewhat hijacked my own thread, and begun to discuss the bolt stress at the top bowl in the bowl assembly. I was doing this so that I could better understand the thrust load, and how it is balanced in the pump.

Now, just to be clear, are you agreeing that thrust is balanced in the pump (thus not transmitted to the foundation), or not?

 

[SteveWag]

I am saying that the design of the foundation must include the downward thrust generated by the hydraulic operating forces generated by the fact that the system is indeed OPEN, as you first suggested. In many cases, this force is much more than the dead load of the pump/motor combination. The impellors are forced downward, stretching the lineshaft, pulling down on the motor bearings, resisted by the motor support, the foundation. I have many installations where a 1½ -inch steel shaft is stretched more that one inch when the pump is rotating. In all of my installations, over a hundred, the motor is of the hollow shaft design and fitted with an auxiliary thrust bearing designed to remove the thrust from the motor rotor bearings. This thrust is passed directly to the pump foundation.
Steve

 

[GaTechTheron]

Steve,

What are your thoughts on the comments from bimr and 1gibson, and the references they provide?

 

[bimr]

I would stop right here and suggest that anyone that thinks there should be a thrust load on the foundation should read through pages 250-259 in the link above.

The OPEN hydraulic system is a novel arguement, but it does not hold water. Refer to the thrust diagram in Fig 3-9 in the link on page 251.

Igor J. Karassik was one of the most authoritative pump engineers in the United States and he has stated that there is no thrust load transmitted to the foundation. End of story.

 

[1gibson]

Unless there is a unique situation where the motor has an external support (for example, pump mounted to first floor, motor mounted to 2nd floor) the thrust is resolved within the pump. Based on lack of details in Steve's description, this very well could be the case and his comments would be accurate.

But for a typical sump mounted pump with motor sitting directly on the nozzlehead, it is not accurate. The thrust is resolved within the pump.

Intuitively, with no *external* forces, the pump can only get as "heavy" as the dry weight, plus the fluid weight it contains. *Internal* forces created by the pump don't have any external reactions.

If you try to push a sailboat with a fan mounted on deck, you aren't going anywhere (despite the fact that the air coming into the fan wasn't initially part of the system.)

The extent of thrust reaction on foundation loading (aside from external motor support scenario) would be if the discharge pipe was not continuous, you would have a nozzle effect and associated reaction force / overturning moment that would tilt the nozzlehead back, away from the discharge flange. This is not the case with continuous discharge piping.

 

[BigInch]

No you are not thinking about the torque load correctly. The motor can have short circuit loads and motors are designed to produce starting torques around 50% more than at operational speeds. Torque is resolved first to resisting moments, in this case shear on the bolts, and then a normal force, tension or compression on the bolts + possibly other force components, if the CG of the motor and pump is not located at the CG of the anchor bolts. Ever hear about the "Right Hand Rule". Point fingers in the direction of rotation and thumb predicts the direction of the resulting force. Torque and resulting forces will be opposite during a trip. Start or short circuit torque loads are often the only real reason that pumps need anchor bolts at all.



What would you be doing, if you knew that you could not fail?

 

[MJCronin]

I believe that Big Inch was correct in his earlier post...

The load on the foundation occurs when the empty pump starts up, fills itself with liquid and attempts to "screw itself" (for lack of a better descriptive term) into the liquid below.

As the pump gets to steady state, the load then goes away.....

IMHO

 

[bimr]

All well and good.

However, Igor J. Karassik points out on page 258 in the book that is linked above that the various forces from startup torque or transient water hammers are not of major significance in calculating the foundation and need not be considered.

 

[GaTechTheron]

Thank you all for your help on this one.