Filter before pump ?

[bigpants]

Hi everyone , hope you're all well during this difficult time. I've been using the lockdown to treat the guzzi ( 1100 sporti) to some shiny new powder coating and now putting her back together. For purely cosmetic reasons I'd like to put the filter before the fuel pump instead of the other normal way round. Does anyone have any idea if this is likely to cause problems ?

[Old Jock]

I've never tired it but the reason for putting the filter after the pump is in case it becomes blocked.

Should that happen the pump suction will collapse the filter and drag everything through the pump and the rest of the system

Blocked filter after the pump leads to low pressure in the system after the filter and higer pressure between the pump and filter. The pump has over pressure valve and recircultion built in, in order that this won't cause damage

Thought about it myself for other than cosmetic reasons, to protect the pump for example, but IMHO it's got the potential to play havoc with the entire system

One thing I did with my 1100, this is a Magni so I don't know if you could do it with the Guzzi frame was put the filter up front beside the pump using a fuel pump bracket off E-Bay. Cut one of the mounting points off and placed the filter in it

[bigpants]

Hmm... Thanks Jock , food for thought there. Ive been through four pumps over the years which I think were all caused by a blocked tank vent making the inside of the tank rust up . So since I've got a chequered history there maybe mine is not the bike to test the swap on. Immaculate looking bike and work btw Jock.

[Tinus89]

In general fluid pump engineering with these kind of pumps (centrifugal), the filter is usually placed behind the pump (downstream). The reason for this is that these pumps prime difficult with the additional restriction of a filter in front of it, leading to cavitation in the pump due to low suction/prime pressure. Eventually this can damage the pump.

Now, the engineering practice above mainly applies to high power, steel made pumps, but most likely also applies to the ones on motorcyles in FI systems.

On the risk of the filter imploding and causing all debris to go through the system:
The pump chucks out way over 3bar, which is regulated by the pressure regulator set at ~3bar. A (partially) blocked filter will create a differential of far more than one bar, increasing the pump pressure.
Placing it in front of the pump and it being blocked, causes a max differential of one bar. This is not sufficient to damage the filter in any way. Changing the filter location does not increase this risk.

[bigpants]

Thanks tinus. So , just to make sure I've understood - placing the filter upstream isn't likely to cause the filter to fail  but may cause cavitation causing the pump to fail ?

[Tinus89]

Yes, that is correct.

[Old Jock]

Tinus is right I stand corrected

I looked at data for my pump a Pierburg, which replaced the Weber original and they state in their literature that installation of a filter can stop pump priming and the pump could run dry and be damaged

Not sure about the Weber or Bosch but the Pierburg is toothed ring gear pump, it looks similar to a centrifugal pump characteristic as pressure rises flow rate reduces

[n3303j]

In general fluid pump engineering with these kind of pumps (centrifugal), the filter is usually placed behind the pump (downstream). The reason for this is that these pumps prime difficult with the additional restriction of a filter in front of it, leading to cavitation in the pump due to low suction/prime pressure. Eventually this can damage the pump.
<SNIP>

Far as I know these pumps are positive displacement pumps (like power steering) not centrifugal (like car water pumps). Positive displacement with a relief valve is the only way they reach the required pressures.

Filter downstream is is a good idea because if the pump starts to disintegrate the filter catches the bits and pieces before they enter the injectors. Got to remember the whole pump motor is running under gasoline (to cool it) and bits will flake off.

[oldbike54]

 Basically it is easier to push liquid than to pull liquid .

 Nice setup Jock , very clean 

 Dusty

[Old Jock]

Thankyou Dusty & bigpants

I always thought these pumps were Centriugal or some derivative of that nature but only to willing to be educated and proved wrong.

Never actually bothered to find out

The system on the Sporti only runs around 35-40psi well within the capabilty of a centrifugal.

Diesels of course are whole different ball game

[elvisboy77]

Hi everyone , hope you're all well during this difficult time. I've been using the lockdown to treat the guzzi ( 1100 sporti) to some shiny new powder coating and now putting her back together. For purely cosmetic reasons I'd like to put the filter before the fuel pump instead of the other normal way round. Does anyone have any idea if this is likely to cause problems ?

Don't do it!  There is a reason it is where it is shown, the reason is cavitation.  You could destroy your fuel pump.  Glad you asked rather than just doing it.

[Wayne Orwig]

Filter first sounds like a recipe for vapor lock, where gasoline vaporizes under suction, and the pump can't pump the fumes.
The spine framed bikes had an issue with that around 2000, before the pump was put in the tank. Putting the pump and pickup right in the fuel fixed that.

[n3303j]

Thankyou Dusty & bigpants

I always thought these pumps were Centriugal or some derivative of that nature but only to willing to be educated and proved wrong.

Never actually bothered to find out

The system on the Sporti only runs around 35-40psi well within the capabilty of a centrifugal.

Diesels of course are whole different ball game

Did a pile of Web looking after my last post. About as many ways to pump fuel as there are designers. Didn't see any centrifugal pumps where input is into the center of a rotor and the spinning centrifugical force propels the liquid out to a peripheral discharge.
But turbine pumps abound where a rotating member runs in a close fitting housing and intake and discharge ports are on the same diameter but opposite sides of the rotating disk. The disk runs between two tapered chambers.
There are also Vickers Vane type pumps, gear pumps and roller vane varieties.

Replaced the pump om my V11 last fall and tossed the old one before I headed out to the Florida winter. Now wish I had saved it so I could cut it open and satisfy my curiosity. Certainty have plenty of spare time at the moment.

[Tusayan]

Net Positive Suction Head (NPSH)...  look it up

Basically you need a certain amount of pressure at the inlet of the pump to make it hold prime, with that pressure requirement increasing with flow through a variable flow pump.  The typical available NPSH on an injected motorcycle fuel pump is the gravity pressure/head of the vertical distance between the fuel level and the pump inlet, minus the pressure drop through any restrictions in the inlet line - which would definitely include the restriction to flow of a filter.

Some pumps in other applications need so much NPSH that they have another pump to feed them, a boost pump on the inlet.  The boost pump doesn’t require much NPSH at all, but has limited outlet pressure.  It might be a centrifugal pump boosting inlet pressure for a high pressure reciprocating pump. The centrifugal boost pump makes enough pressure to prime the high pressure pump and prevent cavitation in the inlet with increased flow.

[n3303j]

1998 V11EV  Bosch 0 580 464 048.
Cleaning workshop I found the old fuel pump.
It is a positive displacement roller pump.

An odd bit of trivia: The shaft through the center of the armature is solidly mounted in the housing. There are bearings inside the armature that let it spin on the static center shaft. There is a plastic coupling between the spinning armature and the pump rotor. Nifty.

[dguzzi]

I would suggest a rock catcher (screen) type filter in front of the pump and leave the fuel filter as is.  That is if you were inclined to do the project after this discussion.
  I agree Jock, very nice set-up!

[Old Jock]

1998 V11EV  Bosch 0 580 464 048.
Cleaning workshop I found the old fuel pump.
It is a positive displacement roller pump.

An odd bit of trivia: The shaft through the center of the armature is solidly mounted in the housing. There are bearings inside the armature that let it spin on the static center shaft. There is a plastic coupling between the spinning armature and the pump rotor. Nifty.

Well pass me a large slice of humble pie 

Thank you very much for going to the trouble

John

[n3303j]

Well pass me a large slice of humble pie 

Thank you very much for going to the trouble

John

There was a pump body diameter ultra fine mesh screen filter at the input end of my pump ahead of all the pump assembly. This catches the big pieces coming down the line. The high pressure filter after the pump catches the pump debris before it can hit the injectors. The brushes, commutator, bearings and pump mechanics wear. The filter after the pump catches these bits.

[lucky phil]

Far as I know these pumps are positive displacement pumps (like power steering) not centrifugal (like car water pumps). Positive displacement with a relief valve is the only way they reach the required pressures.

Filter downstream is is a good idea because if the pump starts to disintegrate the filter catches the bits and pieces before they enter the injectors. Got to remember the whole pump motor is running under gasoline (to cool it) and bits will flake off.

Yep, basic hydraulic system engineering. The pump is the wear/failure point and the filter is there to protect the system from wear particles and debris from a pump failure.

Ciao

[Kiwi_Roy]

Slightly off subject.
On return from a working trip to Niagra I diverted to Denver Colorado where I bought a 89 EV sight unseen, my Wife was away at the time 
I picked up the bike and rode it to Ouray Colorado to attend NAR organized by Rocker59. The first thing I noticed was how noisy the fuel pump was and going up some of the hills it would only make it to the top by dropping a couple of cogs, If I stopped the engine on restarting it would seem ok for a few minutes then bog down again. I was a little worried how I was going to get back to British Columbia as by this stage it would barely make it on the flat let alone any hills so I skipped out of the days events to troubleshoot the bike.
Eventually It occurred to me that the fuel filter may be blocked so I took it off and let the trapped fuel run backwards into a pop bottle, it had a good 2 inches of fine rust coloured mud in the bottom. After that the bike ran like a champ all the way back to BC. Eventually I removed the filter again and gave it another backlash to get the remaining rust out.
Later on Patrick told me how those teardrop tanks with the recessed fuel fillers have a curly pigtail drain that disappears through the tank to come out underneath where it can drip fuel onto the hot engine. Its not supposed to do that of course it's supposed to have a rubber hose to drain the overflow safely to the ground. He told me how this drain rusts closed and how to fix it with short pieces of inner cable spinning in an electric drill and a source of compressed air. It's quite a job to get one unplugged but very satisfying when eventually the air is able to blow out a bunch of rusty mud.
I concluded that any water from rain or washing that gets in around the recessed cap sits there waiting until next time you fill up to drop unseen and lay in the bottom of the tank turning the steel to rust, the rollers of the pump make a great grinding mill turning the rust to talcum powder, no problem for a plastic tank of course. If you own an EV or Jackal with the teardrop tank its a dollar to a donut the drain is plugged, especially if it has spent any time outside. Put a few drops of fine oil into the drain to see if it disappears, if not enjoy spending a few hours trying to get it free.
BTW, you can test the pump by substituting your meter on Amps in place of the secondary injection relay, it should draw just over 4 Amps I believe but will be significantly higher with a plugged filter.
https://www.dpguzzi.com/efiman.pdf
Just don't take any notice of how they show the pressure regulator, thats wrong the vent is open to atmosphere.

[bigpants]

Just a slightly late update. I took the advice and left all well alone but I did get a new pump. But this time I bought the very cheapest one I could find , from a Renault kangoo. Primes very quickly but seems to work fine and only £15. We'll  see how long it lasts.