What's new
Van's Air Force

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

Carburetor Problem? Gasoline/Fuel Smell During Test Flight

MartySantic

Well Known Member
Last week before leaving to witness the space shuttle launch, I flew the RV-12 after replacing the prop hub and adjusting the prop pitch. After climbout, I flew for about a minute at WOT (5600 rpm). After reducing the power I noticed a gasoline smell which lingered and it coincided with what felt like a momentary engine miss or shudder. I immediately reduced power and headed back to the airport. The same happened on my return to the airport.

After landing I immediately pulled the top cowl. The was ZERO evidence. No gasoline smell in the cockpit, in the engine compartment, the air filter, the fuel pump, or the carburetor overflow tube. The carb tray was dusty and there was no evidence of gas spilling onto the tray and creating a track in the dust.

With assistance, started the engine and ran the engine (on the ground) at various power settings to WOT for about 20-30 seconds with a friend observing the engine. Zero evidence of a fuel leak at the carb, the overflow tube, the fuel pump or any of the connections/fittings. There was no gasoline smell. No evidence of an engine miss on the ground. The gascolator sample was clean before the test flight.

Help! I am looking for troubleshooting ideas!! Based on a couple of previous threads.....carb float problem (but, it self corrected)? speck of dirt in the carb needle valve?

Will pull the bottom cowl today and continue to look. Hesitant to fly until I find something.
 
Last edited:
I have found that ALL aircraft engines seem to shudder when I smell fuel :D
With the tank and many feet of fuel supply and return lines going thru the cockpit not to mention the fuel pump, shutoff valve, and fuel flow sensor, it may be prudent to pull some inside panels and look for evidence of leaking there as well.
 
More info requested!

What were the indications on your EMS? Any alarms going off. Loss of Fuel Pressure or anything else? If your don't remember then try down loading your Dynon on to computer to check. If you have a long runway I would try another flight if you cant reproduce it on the ground. On takeoff if you get good power climb at Vx to altitude and circle the airport. if you experience a loss of engine power at full throttle and an indication of a loss of Fuel Pressure then my money is on the fuel pump. It can fail with such a small leak that by the time you check it could have evaporated. How many hours on engine. What kind of fuel? Did the engine run smoother at lower power settings?
 
I have found that ALL aircraft engines seem to shudder when I smell fuel :D
With the tank and many feet of fuel supply and return lines going thru the cockpit not to mention the fuel pump, shutoff valve, and fuel flow sensor, it may be prudent to pull some inside panels and look for evidence of leaking there as well.

If I had a leak in the fuselage, I would expect the smell to linger with the canopy cllosed. No such smell. But will pull panels and check the fittings.

What were the indications on your EMS? Any alarms going off. Loss of Fuel Pressure or anything else? If your don't remember then try down loading your Dynon on to computer to check. If you have a long runway I would try another flight if you cant reproduce it on the ground. On takeoff if you get good power climb at Vx to altitude and circle the airport. if you experience a loss of engine power at full throttle and an indication of a loss of Fuel Pressure then my money is on the fuel pump. It can fail with such a small leak that by the time you check it could have evaporated. How many hours on engine. What kind of fuel? Did the engine run smoother at lower power settings?

There were no alarms on the Dynon. All parameters seemed to be normal. On takeoff had a good climb at Vy, 5150 rpm. 100 hours, using MOGAS. I will download the Dynon log. Can a fuel pump problem be intermittent and self correcting? The shudder was momentary and it is hard to say if reducing the throttle helped matters.
 
Last edited:
Get that problem to reproduce!

If your engine runs ruff at full power but when throttled back it smooths out and runs fine that could be a fuel pump. You would also see a loss of fuel pressure and an increase in fuel flow to the red line. You need to figure out a way to try to reproduce the problem. Until then we are just guessing.
 
smell

I know that there are obvious differences between the RV12 and the RV8, but this is what I found when the smell of gasoline scared me to death.

The fuel tank vent line openings are directly beneath the pilots seat on the underside of the fuselage. At different times I would catch a whiff of fuel. I found out that the only time I ever smelled it was when both tank were packed full and I made a turn in one direction or the other. Fuel would leak from the high tank vent opening and I could smell it inside the cockpit even though I was moving through the air at cruise speed. It never does it unless the tanks are very close to full. These vent line openings are facing forward and are somewhat "pressurized" by the air as the airplane flies through the air. Evidently the pressure in the tank is greater when the fuel is aggitated and there is very little air in the high tank. Just my experience for what it's worth.
 
Data Download?

Marty, if you haven't already, pull the data from the Dynon and look closely at the fuel pressure and fuel flow at the time of the incident. If your engine-driven pump dumped fuel (the odor) it will be reflected in data. (FP down, FF up.) Unfortunately car gas will not leave distinctive blue stains on the weep holes on the pump. If you swab the weep holes carefully, you may be able to pick up a yellow stain. You might also check the carb vents (clear hoses) in case some kind of over-pressure caused them to barf fuel.

FWIW, Rotax tech told me authoritatively that pump does *not* have an intermittant failure mode.

Jim
RV12 flying 30 hours
 
I don't have anything to add to the good advice you have already received, but the fact that the electric fuel pump is energized all the time has always struck me as odd. Does anyone know if the cert. aircraft that use the Rotax have this full time redundancy? As most know, the use of aux. elect. fuel pumps is nothing new and very effective. The mechanical fuel pump on the Rotax seems to be suspicious at the very least. Good luck, if anyone will figure this out, I know you will.
 
Vapor Lock ?

Marty,
What was the outside air temperature? A RV-12 owner in Texas experienced vapor lock shortly after takeoff. He was burning auto fuel also. His engine sputtered at full throttle but ran fine at reduced throttle. When it is hot inside of the cowl, the fuel can vaporize inside of the fuel lines between the gascolator and engine driven fuel pump. That gasoline vapor can escape out of the carburetor vents without leaving a trace in the drip pan. A shroud around the gascolator along with a scat tube supplying cooling air will help. I am not saying your fuel problem is due to vapor lock, only that it is a possibility. If the Dynon data log shows a drop in fuel pressure and an increase of fuel flow at the time of the problem, then suspect either vapor lock or a fuel leak.
Here is an idea that I had but have never tried: A very small piece of Styrofoam placed under the engine fuel pump weep holes will melt if gasoline drips onto it.
The gascolator screen should be checked also.
Joe Gores
 
Last edited:
Downloaded the data from the Dynon. Something was most definitely going on with fuel flow and fuel pressure. I have attached 4 plots. The 1st is from a month ago. The 2nd is a test flight where I experienced no problems. The 3rd is the test flight where I experienced the gas smell. And the 4th is from the ground run after removing the top cowl, immediately after landing in an attempt to resolve the problem. No gas smell was observed during the ground run.

GtSx-R93x9vq6Lo7ked2NqgEuBuVHOUxKN_qzz5TSVc


U9C1Ax1hcuAi6mZhyw44-qgEuBuVHOUxKN_qzz5TSVc


D65qt0G2-WSDLqvtAIh5tagEuBuVHOUxKN_qzz5TSVc


https://picasaweb.google.com/lh/photo/oph3OjT4HE-oAhWVJRyBAqgEuBuVHOUxKN_qzz5TSVc?feat=directlink

Any insight?
 
Last edited:
Styrofoam

Joe, Great idea! I was going to tape a paper towel fragment in there. Styrofoam is perfect.

Jim
 
Check your fuel pump weep holes. It may only be a little discoloration (brown) in the weep hole rim. Any discoloration is a pump failure.
 
Wow, that is awesome information Marty, I had no idea that data was available to us in that form. Do we assume the left scale is rpm and the right is gallons per hour and/or pressure? The bottom scale defies my guesswork.

Added a fourth plot to the previous post showing the fuel pressures and flows from a month ago.
 
Last edited:
More pump info!

Marty I didnt down load the Data from that Fuel Pump event because we were able to duplicate the symptoms over and over again. We have a 6000 foot runway at our airport and we were able to run down the runway at full power and duplicate the problem then abort the takeoff. At max RPM the engine would start to run ruff, then Fuel flow would cause the first alarm when it would reach the Max red line, then we noticed that fuel pressure would star to enter the yellow arc at the same time just slower than the fuel flow. This happened each time we ran down the runway at full power. We did several full power Static runups at the hanger and were unable to duplicate the problem but when we ran down the runway as speed increased RPM increased enough to cause the problem. Also because of the very small amount of fuel leaking out those weep holes that confirmed to us it was the pump that was bad. Just recently on the Rotax-owner.com web site they added a video about fuel pump failures which said if there is any leakage at all then the pump is bad. Thats why it would be nice if you could duplicate your problem or find a leak in the pump. Your pump maybe just barely failing at high RPM only. Our engine would run fine at a lower power settings. It only occurred at Max power settings. I figured you might be out of warranty by now so I know you don't want to spend a lot of money on a part that might not fix your problem. I dont think you are getting vapor lock. Cause of the power setting you were at when you started having your problem I figured you fuel flow was around 5.4 GPH or 1.3 Gals every 15 minutes which is a fair amount of fuel moving through you gascolator assuming your screen is nice and clean. ;)
 
Wow, that is awesome information Marty, I had no idea that data was available to us in that form. Do we assume the left scale is rpm and the right is gallons per hour and/or pressure? The bottom scale defies my guesswork.

The data is downloaded from the Dynon in the .csv format which you can load into Excel. Using Excel you can look at all of the tabular data or graph it. The left scale is rpm, the right, fuel flow and fuel pressure. The x-axis is time is seconds from engine start.
 
Pulled the bottom cowl and gave the engine a good inspection. No issues. No staining at the 4 weep holes on the fuel pump. All connections tight.

Also called LEAF and spoke to Brett. Has seen my issue in other installations. Indicated a possible cause could be the gas. Gasoline with the winter additives can cause cavitation at the fuel pump with the warmer weather. The cavitation would likely force fuel vapor out of the carb overflow tubes.

As such I drained 15 gal of gas and added a fresh batch of ethanol free premium.

Also pulled the gascolator and replaced the screen. Was pristine.

Weather looks poor for the next few days. Next good day, will do a few ground runs and if all goes well, will fly it and see if the problem recurs.

The one item I see in the data is the fuel pressure is lower now compared to earlier flights. A fuel pump in the early stages of failure?
 
Last edited:
Good Game Plan!

I like your game plan. Fuel could be the problem. We run 91 Octane with no ethanol. We fly about 10 hrs per month here in North Texas. Cavitation in the pump now that is interesting. You might want to try a few gallons of 100LL, I dont think they add any winter additives to it.:confused: Stay close to the home field and let us know how it goes.
 
Would appreciate the following data from all,

With the engine off and electric fuel pump on write down the fuel pressure.

At 4000 rpm during a runup, write down the fuel pressure and fuel flow on the ground with both pumps, then pull the fuel pump fuse and write down the pressure and flow with only the engine driven pump. Send this as soon as you can.

Would like to see your ground data and compare it to my data before I fly.

Then the next time you fly, At 5000 rpm, write down the fuel pressure and fuel flow in the air with both pumps, then pull the fuse and write down the pressure and flow with only the engine driven pump.

Please post your results.
 
Last edited:
Data

Flew today. Hadn't seen your post but I've been watching my fuel numbers closely for past 2 months and pulled my download:

Engine off, Facet on: 1.3 PSI

After landing, 2,500 RPM: 3.7 PSI, pull boost fuse, 3.7 PSI

Fuel pressure @5,000 RPM: 4.1 PSI (I don't pull the boost fuse in flight after what I have been through!) On one hour flight, FP did not drop below 4.0 at any time, RPMs 5100-4500. Prior to takeoff and after landing FP dropped to 3.7. This doesn't make sense to me but I will take the in-flight numbers any time.

OAT was 92 on the ground, car gas.

30 hours flying time, blast tubes to gascolator and fuel pump.

Oh yes, I tie-wrapped small pieces of styrofoam to the FP weep holes: no seepage observed upon removal.

Jim #264

P.S. I tested fuel on styrofoam prior to flight. Avgas does *not* affect styrofoam; car gas eats it up.
 
Quite interesting Jim. I guess the part that got my attention the most was about 100LL not eating styrofoam! Would have never guessed that.
 
Marty,

Sorry about your little mystery. I will provide the requested data next time I fly, today or tomorrow. John
 
My numbers..............

Would appreciate the following data from all,

With the engine off and electric fuel pump on write down the fuel pressure.

At 4000 rpm during a runup, write down the fuel pressure and fuel flow on the ground with both pumps, then pull the fuel pump fuse and write down the pressure and flow with only the engine driven pump. Send this as soon as you can.

Would like to see your ground data and compare it to my data before I fly.

Then the next time you fly, At 5000 rpm, write down the fuel pressure and fuel flow in the air with both pumps, then pull the fuse and write down the pressure and flow with only the engine driven pump.

Please post your results.

Marty,

Took the '12 to New Jersey today for a regional CFI meeting. Took down some numbers for you:

Engine off, electric fuel pump on: 1.5PSI, 1.2GPH.
4000RPM runup, 3.4PSI, 4.4GPH with two pumps on.
4000RPM runup, 3.8PSI, with fuse pulled.
(Why higher? Who knows?)
6500', 5300RPM, 22.5", 3.9PSI - 3.6 with fuse pulled.

Hope this helps.

John
 
John, thank you for the data. So far I have made several ground runs and one flight. And, I agree with Joe Gores. The fuel pressure sender is prone to be inaccurate because it has a 35 psi range and we are trying to measure in the range of 3-5 psig. Per the "nuclear industry" instrumentation knowledge base (my former job), attempting to measure anything in the lower 10% of the instruments span is a big no-no.

My data as measured and downloaded from the Dynon on the last ground run was,

0 RPM, Electric Pump On - 1.2 psig
2250 RPM, Both Pumps On - 3.56 psig, 1.94 gal/hr fuel flow
2250 RPM, Electric Pump Off - 3.56 psig (no change), 1.94 gal/hr fuel flow

When I pulled the fuse, the pressure remained constant but the fuel flow dropped, 1.94 gph to 1.50 gph and then recovered. When I reinserted the fuse, the fuel pressure remained constant but the fuel flow increased, 1.94 gph to 2.56 gph and then recovered back to the 1.94 gph value.

Don't ask me why. The fuel pressure was a straight line on the plotted data. (It may be a little too constant).

The same thing happened during the ground runup at 4000 rpm. Fuel pressure remainedd constant but the fuel flows changed and recovered when I pulled and reinserted the fuse.

The flight went fine. Looking at the Dynon data, the fuel pressure and flows are relatively constant, and change as expected with rpm. No wild variation in fuel pressure or flow as observed in the Dynon download from my problem flight.

It may have been the winter gas and vapor lock. If the problem reoccurs, will order one of the NEW fuel pumps.
 
Last edited:
My numbers, Marty

I think this is really gonna confuse things. BTW, my K factor is set to 96500 since that has some impact.

2500 rpm, on ground
Both pumps on: 2.1 gph 4.1 psi
fuse pulled: 1.7-1.8 gph 4.4 dropped to 1.7 then back to 2.2

4000 rpm on ground

Both on: 3.1 gph 3.8-4.1 psi
fuse pulled: 3.1, dropped to 0.4, back to 3.7gph 1.4-2.3 psi

in both cases, when the fuse was re-inserted, fuel flow readings increased enough to cause an alarm and then dropped back to normal.

I DON'T understand the big drop, then recovery in fuel flow readings since the pressure doesn't follow. In light of the above I'm going to try the styrofoam check.

From whom do you buy the new pumps and I assume the bill is sent to Rotax in Canada.

Wayne 120241 N143WM
 
Hi Everybody,

My fuel pump seems to be one of the "good" ones. See pressure numbers in earlier post. However, I am thinking about replacing it with the new improved model anyway on general principles. Does anyone know what the cost of the new pump is, and how you would order one?

John
 
No problems after another hour of flying today. No fuel pressure or fuel flow alarms. It may have been the winter formulated (alcohol free) premium gas during my problem flight with the OAT in the upper 80's.
 
Using 100LL exclusively. No problems so far. I installed a quickdrain to assist with those more frequent oil changes.
 
Mitch Locke is running 100 LL exclusively. I think he uses a product to scavenge the lead and does more frequent oil changes. I don't think he has had any problems.
 
Back
Top