Yes
In fact it is easier to get your before and after straight pipe distances on the suction side for mag meters which typically recomend 3d to 5d up and 1d down. You will need that short spool piece between (at the least) due to the high swirling action of the impellor.
But if the meter is a high pressure drop mechanical device (such as a disk or piston at high flow) it is not recomended unless available NPSH can cover the pressure drop
You should also try flow straighteners in case if you want to place your meter in the pump suction. Conceptually, it's not a bad idea if you have enough NPSHa after meter installation, like hydrae pointed out.
Actually, we loose the flexibility of choosing smaller meters when they are placed in the pump suction. A flowmeter can be selected for a higher velocity than the pipes are designed for, to save the cost(I did it for even 6m/s when the pipe was designed for 2m/s)
As the pump discharge velocities are always higher than that of suction, the meter size becomes inherently lower. It all depends upon what is the acceptable pressure drop.
I also agree with Hydrae that adding friction to the pump suction can be dangerous. It all depends on your situation. You need to verify the NPSH requirements.
You can use low headloss meters such as mag meters, ultra sonic meter (depending on your liquid) to measure at the suction but, depending on your accuracy requirements, remember the run lengths before and after the meter.
Upstream flow restrictions, of course, subtract from the NSPH available (NPSHA) for the pump. But there's another consideration -- fluid energy (and head is simply a matter of the energy available per pound of fluid) is in short supply at nearly all pump suctions. And simply keeping the NPSHA above the pumps NPSHR (required) is not sufficient -- NPSHR, quoted where the head falls off in a pump suppression test, is where the pump inlet is so blocked with vapor that the pump suction is effectively occluded. Well before this, cavitation exists, and its potential for damage is there.
In tests I performed in the 80's with suction viewports, to supress cavitation took NPSH levels 3 times that at head fall-off.
Further, in the discussion of flow straightening, without agressive flow straightening techniques (honeycombs, etc.) typical length requirements are closer to 40 diameters of upstream pipe. I have performed tests with flow-traversing pitot tubes, where agressive flow straightening vanes and honeycombs were used (.18 in cells in a 12 inch inlet), and it 5D was nowhere near enough to smooth out the velocity profile.
So, in short, if you can avoid it, put the meter at the discharge where fluid energy is high instead of at the suction where energy is in short supply -- the risk of less than satisfactory suction performance and/or cavitation damage is greater there. But simply doing some pressure drop calcs and assuming that because NPSHA is greater than NPSHR, everything is OK could give you a nasty surprise.
The placement of flowmeters in the suction lines of centrifugal pumps is not a good idea. Apart from the sound points about the need for straight lengths of pipe and the size of meter that can be used there are others.
In the suction line there can be fluid rotation coaused by the pump particulalry at start up, but even after.
Another consideration is that the larger meter to suit the suction conditions not only costs more but is operated in the lower part of its range. The meter may not be a s accurate at the % of its span.
