What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

90 Degree Bends and Fuel Cavitation

rlmccarter

Active Member
Hi Guys,

Have any of you experienced cavitation (aka vaporization) due to 90 degree fittings on the supply side of your pump?

I'm putting an SDS injection system on the -8 and just bolted in the beautiful twin pumps. They are a tight fit so I have 90 degree fittings on intake and output. To make it interesting, I'm also using an Andair valve with 90 degree fittings. The tanks of course have 90 degree fittings coming out of them.

So tracing the fuel from the tank it will go through five 90 degree bends: at the tank exit, the valve entrance, in the valve, the valve exit, and the fitting into the pump.

I can eliminate three of the five bends by drilling four more holes in the floorboard to reorient the pump and getting straight connectors into the Andair. I might want to use mogas in the future, but this is just a four cylinder engine. The fuel flow can't be that high. Am I obsessing?

DSC_6076-M.jpg


Cheers,
Bob
 
Am I obsessing?..

Nope.

Anything you can do to reduce the number of 90's on the supply side is a good thing.

Remember that even at idle, the pump is going to be pushing a lot of fuel through the bypass regulator so the fuel flow (vs. consumption) is going to be high. Get the pump as low as possible relative to the fuel tanks and as close as you can to them. Im not a fan of the stock -8 pump location, but its a lot of work to move it.

My SDS pump module is in the stick well on my Rocket with "zero" 90's on the suction side. A major plumbing headache, but worth it to me.

The -8 and Rocket are quite different, but my pictures are found here:

http://www.homebuiltairplanes.com/f...t=26077&page=2&p=392515&viewfull=1#post392515

Might give you some ideas
 
Last edited:
Also... After looking at the picture it appears you have the module backwards. (The inlet filter is on the firewall side). Am I seeing something funny?
 
Rerouting begins!

Hey Michael,

Good points.

Yup I had the inlet toward the front as an intentional part of the fuel tubing plan.

Based on this I'm changing plans to mount the pump about an inch above the floor on brackets and orienting the inlet aft. That will give straight routing into the pump. I will also put straight AN-6 connectors on the Andair supply side inlets.

Thanks for the sanity check and quick reply!

Bob
 
Also, dont worry about a 90 at the outlet if it helps make the strain relief loop to the firewall - at 50 PSI it wont even notice.
 
Consider mounting the pump vertically on the gear tower forward side. This allows for smooth loops to/from the pump, and provides some flex tolerance. I try to avoid short straight rigid tubing hard mounted at both ends.

Carl
 
Vertical mounting raises the pump inlets even higher than they are now. You want them as low as possible relative to the fuel.
 
Vertical mounting raises the pump inlets even higher than they are now. You want them as low as possible relative to the fuel.

Pump lift is an issue - however on a vertical pump with inlet at the bottom and outlet looped back down to inlet height it would have just the fluid flow drag of the pipe itself - about the same if the pump was horizontally mounted.

Carl
 
A valid perspective if you consider the inlet of the pump "module" synonymous with the impeller location. I could be wrong, but I think of the impeller plane as the salient "inlet" WRT fuel level. And in a vertical mounting, the impeller plane of the pump is going to be notably higher than if horizontal.
 
We recommend the pump module be mounted horizontally, as low as possible in the airframe on low winged aircraft like the RV with wing tanks. Minimize the use of 90 degree fittings on the inlet side of the pumps.

Tens of thousands of flight hours with this layout shows it to be the most reliable.

We don't recommend the pumps be mounted firewall forward.

Whatever you do, don't mount the module vertically with the inlet facing up. This has caused cavitation and short pump life in at least 2 instances.
 
Sort of related:

Ross, are most people drilling a bypass orifice in their fuel pressure regulator? Mine is equipped with one that was about the smallest drill but I could find. I notice when I first turn on the pumps, they growl a little bit for a second or two while the system purges vapor through the returns, then they are fine. 700hrs on the airframe with this fuel pump setup without a single problem.

my pumps are mounted horizontally on the floor in my -7a.

For what its worth, my system has several -6 90's in the suction side, as well as a large area, prefilter.

Caleb
 
Sort of related:

Ross, are most people drilling a bypass orifice in their fuel pressure regulator? Mine is equipped with one that was about the smallest drill but I could find. I notice when I first turn on the pumps, they growl a little bit for a second or two while the system purges vapor through the returns, then they are fine. 700hrs on the airframe with this fuel pump setup without a single problem.

my pumps are mounted horizontally on the floor in my -7a.

For what its worth, my system has several -6 90's in the suction side, as well as a large area, prefilter.

Caleb

There is no need for a bypass hole in the regulator as the valve already bypasses as much fuel as is required to control the pressure vs. engine demand. I never understood why Jan did this, it was not a fix for having the pumps in a hot place or poor orientation on some installations.

The recommendation to avoid 90s is best, safest possible practice. Many people have at least one 90 before the pump at the fuel selector. The idea is to reduce restriction to minimize the chance of pump cavitation to a minimum. We want the pump inlets flooded so they never need to pull fuel.

The problem is not with 100LL at sea level on a cold day, it's running winter blend mogas on a warm day at high altitude where even the slightest inlet restriction can cause the fuel to "gas".
 
...The problem is not with 100LL at sea level on a cold day, it's running winter blend mogas on a warm day at high altitude where even the slightest inlet restriction can cause the fuel to "gas".

Which explains the lengths I'm going through to minimize inlet restriction - summer in Mojave + big fuel flow + high altitude + "California blend" auto fuel = big potential for vaporization.

My attempt to run that foul car gas may not work out, but I'd hate to leave a 90 in the system and always wonder if that was the one tiny factor that tipped the scales out of my favor. At least now if it vaporizes, I know I did everything I could to prevent it.
 
Last edited:
Do we know first hand of someone who has had vapor issues with a high pressure fuel system? I flew my plane to 115F+ Redding CA a few times while running 87 E10 in the Subaru. This was done while using the same fuel system as now. In my case the only difference I saw was maybe an extra second to prime the system when I first turned on the pumps. I'm certainly not saying it isn't of concern, Just curious.

Caleb
 
Do we know first hand of someone who has had vapor issues with a high pressure fuel system? I flew my plane to 115F+ Redding CA a few times while running 87 E10 in the Subaru. This was done while using the same fuel system as now. In my case the only difference I saw was maybe an extra second to prime the system when I first turned on the pumps. I'm certainly not saying it isn't of concern, Just curious.

Caleb

Yes, we've seen this several times with poor pump placement on mogas in the summer. We find the engine often continues to run ok but the pumps can fail in a few hours time. They don't take kindly to sucking a vapor and fuel mixture.

Your installation seems to have plenty of margin. That's good that you've tested at the extreme temps. Yikes. You fly at 115F?!!!

The most important thing is to have the inlet as low as possible on an RV which you do.
 
Last edited:
Hi Guys,

I'm aiming for a simple design with margin for worst case (mogas at high ambient temp and altitude) - may never do that, but I like margin.

In the -8 there is enough room for SDS' mounting recommendation (low, horizontal, straight inlet) with 90 degree outlet and a service loop to the firewall.

My tech counselor is dropping by tomorrow to put a second set of eyes on the design before I finalize. I'll post pictures in a few days in case it helps others.

Cheers,
Bob
 
pump(s) in tank

One of you clever engineers/businessmen need to design a kit that puts the pump in the tanks where they belong. Either one in each tank, two in each tank, two in one tank plus a transfer pump. Come on, guys! :)
 
Hi Guys,

I'm aiming for a simple design with margin for worst case (mogas at high ambient temp and altitude) - may never do that, but I like margin.

In the -8 there is enough room for SDS' mounting recommendation (low, horizontal, straight inlet) with 90 degree outlet and a service loop to the firewall.

My tech counselor is dropping by tomorrow to put a second set of eyes on the design before I finalize. I'll post pictures in a few days in case it helps others.

Cheers,
Bob

You might look at angling the pump module to give you more room for fittings in the diagonal. You can also discharge from one of the side fittings and plug the end fitting to gain more room on the inlet end. Don't fuss too much. One 90 on the inlet should be just fine with the pumps on the floor. Very few people get away with no 90s and they are all working fine.

We're only saying best practice is to avoid putting in multiple 90s on the inlet side.
 
One of you clever engineers/businessmen need to design a kit that puts the pump in the tanks where they belong. Either one in each tank, two in each tank, two in one tank plus a transfer pump. Come on, guys! :)

More complexity with more chances for failure. The recommended setup is proven with tens of thousands of successful flight hours. No need to re-invent the wheel and experiment here IMO.
 
I do agree with Ross on the in tank pump location. We have done a few in the wing roots, that helps the situation some. Plumping is a pain.

Tom
 
Do we know first hand of someone who has had vapor issues with a high pressure fuel system? I flew my plane to 115F+ Redding CA a few times while running 87 E10 in the Subaru. This was done while using the same fuel system as now. In my case the only difference I saw was maybe an extra second to prime the system when I first turned on the pumps. I'm certainly not saying it isn't of concern, Just curious.

Caleb

Prior to cleaning up my inlet-side plumbing, I had a couple instances of a pump refusing to pick up its prime while cruising at 16,500 and 17,500 while running 91E10. I run dual pumps, one pulling from each tank, so running on one tank at altitude for a while allowing the fuel in the other line to slowly vaporize in the lower ambient pressure at altitude and then when it was time to switch tanks it wouldn't pick up the prime to pump.
 
Hi Guys,

I'm aiming for a simple design with margin for worst case (mogas at high ambient temp and altitude) - may never do that, but I like margin.

In the -8 there is enough room for SDS' mounting recommendation (low, horizontal, straight inlet) with 90 degree outlet and a service loop to the firewall.

My tech counselor is dropping by tomorrow to put a second set of eyes on the design before I finalize. I'll post pictures in a few days in case it helps others.

Cheers,
Bob


How about a straight outlet off one of the side ports (aiming toward the aircraft centerline) with a 270 loop to a 90 degree bulkhead fitting at the firewall?

Its been a while since I plumbed the -8 (and I moved the firewall fitting to the center anyway).
 
Your installation seems to have plenty of margin. That's good that you've tested at the extreme temps. Yikes. You fly at 115F?!!!

Vapor pressure of mogas blended for Canada and northern California may be different, just as winter and summer blends are different.
 
I must first admit that I haven't built an -8 so I can't visualize the challenges of pump location vis a vis the firewall. I'm reading in several comments here the need for a 90 degree fitting and a service loop on the outlet between pump and firewall.

Would this perhaps be a situation where a stainless braided conductive Teflon flexible line might be the best solution? It would perhaps eliminate the need for a service loop. I'm also thinking if clearance is the issue driving the need for a 90 on the outlet of the pump, perhaps a "swept 90" fitting on the end of the aforementioned flex hose might be a better option to reduce flow restriction?

I know aluminum hard line is cheap, light and often highly reliable. This particular locale might be one where adding three or four ounces of weight might bring significant payback in improved performance and ease of maintenance.

Again, I have no direct experience in this exact installation so my question is an opportunity to learn more about the peculiarities of this installation.
 
You can see in the first photo in this thread that the pump module has 1/4 NPT plugs on the sides. If fore/ aft clearance is an issue, I'd just plug the center outlet and use one of the side ports for the discharge. As Toobuilder mentioned, the pump is reversed in the photo. The Larger hex fitting houses an inlet filter.

Dan is correct, mogas formulations vary widely and so does their vapor pressure. Our goal with plumbing is to minimize the chances of so called "vapor lock" before the pump inlet by following best practice.

It should be noted that the ID of a -6 90 degree fitting is much larger than the bore of the pump spigots. I don't believe one 90 before the pump would have any impact of flow restriction if the pumps are floor mounted. Adding a couple more possibly would. Be aware that these pumps flow around 40 gph at all times.
 
Last edited:
Would a pair of gradual bent 180s cause a problem, in your experience? I'm thinking of fitting fuel filters in the wing roots and the only way I can figure on doing that is with some bent 180 degree tubes. Don't think I'll need hard 90s anywhere though.
 
If you do need some custom braided hose assemblies to plumb your pumps, Tom is the best source for these.

I think Goodridge came out with some braided hose which you could do very tight bends on with no little or no collapse of the hose cross section 10+ years ago. Amazing stuff. Maybe Tom could comment on that.

Braided hose is a lot easier to route than rigid line and you have less concerns about vibration induced failures over the long term.
 
Last edited:
Just for the sake of discussion---

there is a MAJOR Aerospace fitting manufacturer that makes block or forged style angled fittings for all of their crimped teflon hoses from -3 to -8. This manufacturer is very well known, and has this type of hose ends flying on litterally millions of hoses worldwide. So, I'd defer to them to find out the actual flow loss of their forged style fitting, vs a bent tube design, thats used.

Gee---maybe we'll take a couple of fittings up to AirFlow Performance, and have Don flow them so we'll have real data.

Tom
 
For future reference, here is my final SDS pump installation. Mount the pump on 3/4" channel, and it lines up with gear tower's lower fuel line hole.

That eliminates bends on the inlet side of the pump for a clean installation. Thanks all for the comments.

hgko_IneuxV0RpbzwofivXXZTsUZs8XdCxzaCiyWawuWOSqGbComw1cznMH6YOWh-s3aNE6C19nNaSL8gCmAhf--FZZ6RFdMp_H-RoEEPpxa7CCdCxvrKD1fjSD3lXUkQdKMQnotL-z9MzRnV0WprnVx3pCEmiEto00UUFb5c8r08qWdwDgdBvTX1Ykcw4sLuYCFXJ_V0osdkowbDXxm3aPHnOQsmv0I-_anEAMeCVOrFdjoUGJn6OHnm5dJXatoGGiDJCu4A9A5dkpu9SosxMoyVbPYo_Y5O9ItNS0MHW8MG62no2h2jFzzq_gLnJ7z0MrDmdgYzcROZdop10P02sVD5xV12_6FAE1J9HccAXBhj__5oEjRcXmFAyGq3HPc6X_YjiMLNpUoWyCd6KZ68Fk_yPqWZMPixstBACwmfRZJ_JgJgv3i6AfB-IGy37Cnd7Lru8v8SqY7njF6yDDl1BsebyxDqA7Ns8mDAhfbfFLrrHEc16Q09t7EDNbUYWBvg_3LsldXUcn6-NUwAqtGaPZPN-fFPKgHIK6KpI4RggawriLvslTjhr_ZS6lZK89V76zuU1D5Pk0avClhv0H7athFJkeOR_uBfNtrvm8i_Q3cTJNf0mLY_-ndGx8XxrXy72CStHJoTYjvGWKsB5WMgHfxKwW0uLp8RIzdDcNXfmzt2Q=w9000-h600-no


Bob
 
maintenance

For future reference, here is my final SDS pump installation. Mount the pump on 3/4" channel, and it lines up with gear tower's lower fuel line hole.

That eliminates bends on the inlet side of the pump for a clean installation. Thanks all for the comments.

hgko_IneuxV0RpbzwofivXXZTsUZs8XdCxzaCiyWawuWOSqGbComw1cznMH6YOWh-s3aNE6C19nNaSL8gCmAhf--FZZ6RFdMp_H-RoEEPpxa7CCdCxvrKD1fjSD3lXUkQdKMQnotL-z9MzRnV0WprnVx3pCEmiEto00UUFb5c8r08qWdwDgdBvTX1Ykcw4sLuYCFXJ_V0osdkowbDXxm3aPHnOQsmv0I-_anEAMeCVOrFdjoUGJn6OHnm5dJXatoGGiDJCu4A9A5dkpu9SosxMoyVbPYo_Y5O9ItNS0MHW8MG62no2h2jFzzq_gLnJ7z0MrDmdgYzcROZdop10P02sVD5xV12_6FAE1J9HccAXBhj__5oEjRcXmFAyGq3HPc6X_YjiMLNpUoWyCd6KZ68Fk_yPqWZMPixstBACwmfRZJ_JgJgv3i6AfB-IGy37Cnd7Lru8v8SqY7njF6yDDl1BsebyxDqA7Ns8mDAhfbfFLrrHEc16Q09t7EDNbUYWBvg_3LsldXUcn6-NUwAqtGaPZPN-fFPKgHIK6KpI4RggawriLvslTjhr_ZS6lZK89V76zuU1D5Pk0avClhv0H7athFJkeOR_uBfNtrvm8i_Q3cTJNf0mLY_-ndGx8XxrXy72CStHJoTYjvGWKsB5WMgHfxKwW0uLp8RIzdDcNXfmzt2Q=w9000-h600-no


Bob
Hi Bob, Thanks for the updated picture, it looks very clean. Perhaps this is already obvious, but make sure that you can maintain the pump after your skins are installed. I know I've overlooked this small detail once or twice and had to go back and do it again. Regards, Mickey
 
Back
Top