Actuation Torque Vs Pressure drop curve for BALL VALVE

[thought2007]

Industrial expert’s opinion is maximum torque required to actuate ball valve is
“Torque ability to open under large differential pressure and to close under high flow conditions “

My objective is to determine Actuation torque Vs Pressure drop curve (through digital prototyping) using CFD analysis tool.

Design employ trunion mounted ball valve design approach and its stem are supported by bearings
Flow rate: 17.2gpm
Operating pressure: 60psi
Max pressure drop at open condition: 0.5psid

Now my question is how to built boundary conditions for CFD analysis and I carried out using following conditions in flow model to determine pressure drop for various opening angles but results are varying drastically

Upstream condition
17.2gpm Downstream condition
60psi Unknown

Would request you all is this right method to analyze
1. How ball valve is testing is done on test bench to
determine Torque Vs Pressure drop
2. Do I need to give Pressure drop 0.5psi at downstream
conditions as this requirement when ball is open
conditions than analyze pressure drop for various
opening angles
3. Do i need to give Upstream conditions as 60psi and
downstream conditions as Flow rate 17.2gpm than analyze
pressure drop for various opening angles

Your suggestions are well appreciated

Thanks & regards
John

 

[itascot]

thought2007,

I have carried out tests (cyclic under full rated Dp) on valves, using a digital torque transducer between valve stem (shaft) and actuator, however my objectives were to measure and record initial & subsequential changes in required seating, unseating & stroking torque (togehter with valve sealing capabilities changing).
This was many years ago, nowaday's you can use more dedicated equipment (see threads regarding this in recent past

http://www.eng-tips.com/viewthread.cfm?qid=265631&page=2

).
Remember that the actuator will only deliver the required torque to overcome the valves resistant torque requirement, in other words an actuator set to deliver 20.000 Kgm will only deliver a torque value to shaft = to valve resistant torque (this could be very different from catalogue values).
You probaly have drastic changes, because there are so many factors in the equasion:
How long between closing, pressurising, opening of valve ?
Is flow constant ?
Actuator type ?
Actuator instrumentation ?
Is it the same valve you test, or is it same model & make, but not the exact same valve?
Is the fluid in test loop clean.
Is the supply pressure to actuator (if Pneumatic or hydraulic) constant ?
Is the actuator stroking time constant?
I am sure the tests you are doing at moment have already been carried out by valve companies (as they would above all want to reduce required actuator size & consequently reducing actuation costs), but am unaware of any papers published on this subject.

Ciao,

 

[gerhardl]

Several question marks:

Size?
Fluid?

You might well determine a torque curve testing the torque in a test bench, but for what purpose?

How do you propose to determine the safety factor incorporating variation from valve to valve and variation in operation and standstill times, open or closed?

Your torque for start (loosening from seats) and first-time operation might well (after a longer stop with perhaps drying and sticking fluid) be five-fold the measured torque under conditions of fluid-filled valve 'greased and smoothed'.

In this case mechanical variations might anyway well be larger than the influencing force from a throtteled fluid stream.

To your questions:
Any given valve (also a ball valve) will have a given flow value (gpm) at full opening and at given pressure difference (delta P) over the valve.

If you keep the pressure before and after the valve constant, the flow will start at the ball-valve's max, and at a certain point give a traceable throtteling the flow, gradually reducing the flow to zero.

I am not quite sure I catch the reasoning behind your analysis, but I am a bit leaning to you alternative 3, to be the one nearest a practical possible application:

You have a given pressure level upstream, you have a certain size pipeline and a given size and type of on/off valve (your ball valve), and want to find out how the valve behaves at opening and closing at a certain flow. The flow is regulated by a separate regulating valve downstream, where the regulating valve will limit the flow to the max figure mentioned.

Without checking flow figures and torque figures it seems to me two problematic areas:

a) create conditions where you have stable flow conditions, not to be disturbed by the regulating.

b) the torque itself will be a function influenced as stated above by several interacting factors: internal mechanical forces and frictions in the valve, position of valve ball, laminar or turbulent flow, variation from ball-valve to ball-valve, valve construction, internal varying forces over time and from operation to operation (abrasions, sticking, thightness of sealings, old (worn) or new (tight) sealings, internal roughness of surface of ball and sealings, size of valve (for a large-size valve, low speed and low pressure the contributen from flow forces itself will be lower than other varying forces) etc. etc.

It might, as you state, be difficult to get anything else but largely varying results if you are working in a scale where other forces are larger than the ones you want to analyze.

(Remark: I presume you measure torque directly on the stem? No influencing gear or other device that may give variations according to valve position)

 

[thought2007]

Hi,

Thanks very much Itascot & gerhardI
and I am sorry for delay reply as I was on vaccation for two weeks

gerhardI,

Size : 0.875"
Fluid: Ether poly propelene(Red oil)

The purpose of this analysis is I am developing Actuator to open & close the valve using DC Motor. So I need to find out maximum resisting forces from valve(to over come these resisting forces) to select DC motor for actuator design
Some of the contributing factors
1.Friction forces between all rotary elements(
Seat material is Teflon)
2.Assembly construction of rotary elements
3.Torque ability to open under large differential pressure and to close under high flow conditions

1& 2 determined using classical methods but 3rd factor I am unable to determine analytically then carried using CFD analysis to estimate max forces.

But the problem is the determined torque value is very less (5times less) than the other customer valve for the same capacity valve. I unable to justify the results of torque value which I have calculated using analytical methods & CFD analysis results

The value which I determined is coming : 18lbf.in
Same caspacity valve which i compared with Flowserve valve is approx : 60lbf.in (ref 4the page in below link)

http://www.flowserve.com/files/File...owControl/WorcesterControls/WCENSS0001-01.pdf

This variation making me confuse & unable to find out mistake

Thanks & regards
John

 

[itascot]

I still dont see the problem, valve manufacturers give guaranteed maximum torque values at these pressures, hence they are taking many "internal" calcs into consideration, such as surface finish tolerance, dimensional tolerance and once they come up with a global sum of factors that can have effects on valve rotary motion...they slap on a extra 15% to 30% to these figures, to be on safe side.
With a small rack & pignion actuator, you wont be saving much (if anything) on single or double acting by shaving off a couple of Lb-in...not for this size valve/actuator assemlby. I could understand for instance if you were doing all this for anything above 3" #300 where torques start to have adverse effects on actuation pricing and overall dimensions(change fron rack to yoke types).

 

[itascot]

Sorry, just read through my reply....you want to use a DC motor, what accuracy are you getting in your motor/gear mechanism ? (remainder of previous comments regarding valve torques remain the same).