Pump Speed Vs. Torque Characteristic Curve

[ThePunisher]

Hi all,

We are dynamically modelling motor and associated pump in SKM PTW. We ahve obtained the motor characteristic curves with associated inertia.

However, we only managed to obtain the pump inertia and rated torque. We have requested our Mechanical team for the pump speed vs. torque but all they have instead is a pump performance curve that indicates the head and associated flow rates in gallons per minute.

Is it possible to derive the pump speed vs. torque using the pump performance curve? Is the pump speed vs. torque ONLY obtained from the pump manufacturer as an independent curve data?

 

[7anoter4]

You can calculate the required power BHP=(GPM*H*Sp.Gr.)/3960/eff for any rpm
and from T=5252*HP/RPM calculate the torque or :
The torque=Trated*rpm^2/rpmrated^2 at any rpm.

 

[77JQX]

One can get the flow, head, and power for 7anoter4's equation from the centrifugal pump affinity laws: gpm2/gpm1=rpm2/rpm1, h2/h2=(rpm2/rpm1)^2, BHP2/BHP1=(rpm2/rpm1)^3. Assuming you're talking about a centrifugal pump.

 

[Compositepro]

Pump torque is a function of many factors other than speed alone.

 

[PersonalProfile]

The mechanical engineers should be able to provide a system resistance diagram (which can then be related to pump shaft speed, though pump affinity laws).

I assume you are starting DOL and hence can obtain a Torque / Speed curve for the motor.
You either know or can find a relationship between motor and pump speed.
Knowing" alpha = T (motor torque available - pump resistive torque) / I (motor + pump etc inertia) and

and alpha = dw (change in shaft speed) / dt (time step)

You can assume a dw (say 100 th of the final shaft speed - though you need to try refining to ensure you have selected a suitable dw) and step along the "dw's" until the final shaft speed, whilst calculating "dt" along the way. The sum of "dt's" is your starting time.

I have previously evaluated the starting time by computational methods (stepping in Excel). Definite solutions exist (cf computational methods), though for what you are doing it may be easier and accurate enough to do via computational methods.

Regards,
Lyle

 

[ThePunisher]

First of all, thank you for all your replies.

The starting time is computed as a function of several factors: a) motor and pump inertia, b) motor and pump torque vs. speed, c) motor terminal voltage at different currents (during motor acceleration with pump connected)and d) system impendance/voltage drops during acceleration. Some programs like ETAP would perhaps have an built-in typical centrifugal pump torque vs. speed curve and do curve fitting using the pump rated torque and inertia.

Therefore, deriving the speed vs. torque curve is essential.