Perforated plate as a flow conditioner 3

[Yobbo]

Hi,
I am considering to use a flow conditioner in a suction line of a pump in order to equalise the velocity profile in front of the impeller entrance. The need is there, because I am confronted with an ill designed suction line connected to an inline pump. The pump has already a long history of excessive wear in seals and even fatigue phenomena in the axis. Ideally a 5D straight pipe should connect to the suction flansh to achieve the proper flow to reach the impeller, but in this case a 1.5D radius bend has been connected directly to the suction flansh. I suspect that the eddies resulting from this bend may cause the impeller to be unevenly loaded and thus resulting in early damage or malfunctioning. In the Sulzer handbook I found that a perforated plate or more of them in series may be able to condition the flow without the need to satisfy the 5D requirement. Hydraulically I am able to calculate the resistance. But the pressure drop over such a plate will also cause the plate to be subjected to a uniform load, that must be resisted. Are there any formulas available to calculate the necessary plate thickness while respecting the required hole pattern ( diameter and pitch )? Or should I have it calculated with F.E.M.?

 

[1gibson]

You can try to straighten the flow, but even then you will have more flow towards the outer radius than the inner and it won't solve all of your theoretical problems.

Look into turning vanes within the elbow, they should distribute the flow more evenly between inner/outer radius, and they will also serve to straighten the flow.

 

[zdas04]

I was a measurement engineer in the heyday of the development of flow conditioners for gas measurement. I had an opportunity to review a LOT of proprietary data and the following is from that work. I know, liquid isn't gas, but many of the concepts that were developed then translate exactly.

Check the Reynolds Number on the suction piping. If it is below about 20,000 I would look elsewhere, a perforated plate is not going to help because the flow energy is too low for the plate to act on.

With the perforated plate you want to make sure that the open area is at least 70% of the pipe area (but no more than 76%) so that the permanent pressure drop is minimized (you can't get it to zero, but you can minimize it). Also you want 80% of your flow area in the middle 60% of the pipe.

There is a chance that even with the above that you'll drop the suction pressure too much and cavitate the pump. I'd check that arithmetic really closely before I reduced suction pressure.

Typically a flow conditioner is in front of straight pipe, putting it in front of an elbow might be counter productive, but whatever you do, DON'T put in asymmetrical holes to "trick" the flow into the side away from the bend. You really cannot fool Mother Nature and asymmetric plates always set up a swirl.

David Simpson, PE
MuleShoe Engineering

"Belief" is the acceptance of an hypotheses in the absence of data.
"Prejudice" is having an opinion not supported by the preponderance of the data.
"Knowledge" is only found through the accumulation and analysis of data.

 

[rmw]

http://www.chengfluid.com/publications

rmw

 

[BigInch]

How come everyone these days thinks 10D = 5D

If it ain't broke, don't fix it. If it's not safe ... make it that way.

 

[europipe]

Big Inch, maybe because EN standards double the radius of an elbow,
an inheritage of the DIN std.
Why is also a riddle to me

 

[BigInch]

But, correct me if I'm wrong, Straight Pipe "D" is still "D" if the pipe is Straight. Everyone seems to think that they can shorten the straight length before a pump to 5D without considering if mother nature agrees with their philosophy in every situation. If their pump suction is close to NPSHR, an Ell and 5D "ain't never" going to work.

If it ain't broke, don't fix it. If it's not safe ... make it that way.

 

[europipe]

You are right, this needs a redesign of the suction,
but I think they dont agree with that.
When a designer takes 10D in acount, there are a lot of others (proj. engineers, civil idiots) who are preventing that for a few meters space.
And later on there is no possibility to improve the situstion.

 

[JakeR82]

My only help with the perforated plate thickness would be ASME VIII ligament patterns. It has the efficiencies for hole spacing, and I would assume you could apply those efficiencies to the plate thickness calculations. However, you might not be able to.

 

[Yobbo]

Dear respondants,

Thank you for the usefull reactions. To clarify my sources of information I refer to two books, that seem to me as of authoritive nature:

Kreiselpumpen of Johann Gülich
Handbook of hydraulic resistance of Idelchik

The first book requires knowledge of the german language, although I am not sure whether this vook is also available in english. The required straight length of 5D to 8D was mentioned in the first book, where I must admit that there seem to be a lot of if's involved that make the requirement even stricter. This book states that applying guiding fans in elbows will improve the asymmetrical velocity profile dramatically. As for the straight pipe run of 10D I certainly am curious if there is a more solid background for that figure.

The second book presents the use of perforated plates as usefull in order to improve the uniformity of the velocity profile.

I did include some pictures of the books and the quoted pages, as well as two pictures of the pump involved.


Karel Postulart, The Netherlands
Nuon Power Generation

 

[1gibson]

Yes, turning vanes / guide vanes in the elbow.

 

[BigInch]

Yes there are many English texts that do discuss 5D minimums with straighteners, etc., but the engineer has to realize that even those may be ideal situations that are not uniformly applicable in all cases, which most texts do caution. Straighteners also have a pressure loss penalty, so the benefits may not be equally commensurate. With the situation of an ell connected directly to the suction flange, the supposed benefit the offer may be minimal indeed. In this case you have little alternative possibilities except to add the straighteners, so there seems to be little use in calculating pressure drop. The only slight advantage is the fact that you might know beforehand that it won't help. However in any case, you have no choice, so install it and treat it as an experiment. If it helps, it helps. If it doesn't, remove it and try to find out some way to correct the real problem by reconfiguring the suction line instead of trying to doctor the symptoms.

If it ain't broke, don't fix it. If it's not safe ... make it that way.

 

[1gibson]

Maybe it is an issue of terminology, but just to clarify, a flow straightener (in the form of a flat grid, screen, perforated plate, etc) will not solve all the theoretical problems of having an elbow directly on the pump suction.

The flow may become straighter, but you will still have more flow on the outer side of the elbow, and straight or not this can/will still cause suction instability.

 

[Yobbo]

Thank you for looking at my thread. I do realize that the solution, like the one we discussed, is nothing more than a ( perhaps poor attempt to create a better velocity profile at the suction flange. I fear that my hands are tied for the moment I will not be able to thoroughly modify the situation.
One last question though to get it clear for myself. Can you confirm from your own experience my assessment that it is most probable, that the current connection of a sharp curved bend to the suction flange can be a cause for troubles with the pump?

Karel Postulart, The Netherlands
Nuon Power Generation

 

[BigInch]

ABSOLUTELY.

If it ain't broke, don't fix it. If it's not safe ... make it that way.

 

[Yobbo]

Hello Biginch,

Thank you for the reassurance.
As far as I am concerned this thread can be concluded.
I wish you all a good weekend!

Karel Postulart, The Netherlands
Nuon Power Generation

 

[BigInch]

me2. c u Later.

If it ain't broke, don't fix it. If it's not safe ... make it that way.

 

[hacksaw]

excelleant discussion and advise,

would point out that a perforated plate, flat screen or grid cannot remove swirl caused by elbows

 

[Yobbo]

Hello Hacksaw,

I do agree with your statement. Because of that I will apply a combination of a flow conditioner with guiding vanes in the bend and a perforated plate. With that I hope to reduce the swirl as well. As the other repondants pointed out already one cannot expect a fully well conditioned flow towards the impeller with a solution like this, but if there is no room or money to drastically redesign the suction piping it is worth a try. It has a highly experimental content and may not bring all I wish for.

Karel Postulart, The Netherlands
Nuon Power Generation