MooneyMaint

Sign in to follow this  
Followers 0
  • entries
    3
  • comments
    17
  • views
    508

A primer on hard starts with a hot fuel injected engine

DVA

763 views

Vapor lock comes in varying degrees, so a single technique to purge the fuel lines of “gas air” won’t necessarily work. I break it down by short heat soak (about 10 min or less) and long soak (about 30-60min) and yes, there is that charlie foxtrot area in the middle ~15-30 minutes where anything can happen.

This discussion is for a fuel injected engine.  

During a short heat soak the likely culprit will be the upper fuel lines to the distribution device and the injector lines to the cylinders. Note, these fuel lines often sit atop the hot engine, and since heat rises, the relatively small volume of liquid fuel in these lines atomizes quickly and becomes a vapor (gas air). To fire a mixture off in the combustion chamber, there must be an atomized fuel mist suspended in the surrounding air. If the fuel is too atomized as in a vapor, the fuel density won’t be sufficient for firing, hence a hard start, actually due to a overly lean mixture.

To clear a short soak, you need to pressurize the upper fuel lines with just enough fuel to push out the vapor, and not too much that you flood the intake ports.  This is where most pilots get in trouble with hot starts; a hot engine needs far less fuel to start than a cold engine. The theory is to begin the starting sequence for a short heat soak with NO fuel flow and then ADD fuel slowly until it fires off.  Most often, we do it backwards and that makes things exponentially worse.

During a long heat soak, the entire fuel system comes in to play from the fuel tank feed lines to the fuel pumps, to the pump chamber, to the feeder lines, on up. It generally takes longer for these components to heat up and begin to vaporize after sitting than do the upper fuel lines. The clearing technique here is different than a short soak, as you want to (have to) push the vapor out of the pump circuit and that takes time; sometimes a lot of time.

To clear a long soak, you need to purge the entire fuel circuit of vapor, not just the lines going to the injectors. In a long soak, there is a lack of liquid fuel in the pump circuit (because it got hot and vaporized away) ... and a fuel pump, while good at pumping a liquid, is very inefficient at pumping vapor. So you crank and crank and crank and nothing happens because nothing is happening - no fuel is flowing - because the fuel pump is essentially pumping air. In cases of a long soak, using just techniques that will effectively clear a short heat soak will do little to clear the pump circuit, causing a hard start.

Knowing a little about why it’s hard to start a hot engine often makes it easier to find a solution. 

If this works for anyone send money; I have airplane payments to make and my wife and mistresses want jewelry. 

Short Soak: 

  1. Throttle cracked to the point where it would need to be to have about 1000-1200RPM if the engine were running
  2. Boost pump OFF - (Do not use the boost pump or primer at all)
  3. Mixture Full Rich for about 5 seconds then Idle Cut Off
  4. Begin cranking the engine, wait a 2 seconds then slowly (over 5-10 seconds) move the mixture toward rich. Don’t exceed recommended cranking time.
  5. As soon as the engine starts, keep the mixture at about that point, adjust it and throttle for smooth operation. You should ALWAYS run the engine as lean as possible on the ground.

Rationale: The throttle is cracked open so that when the engine fires, there is proper air flow for the fuel that is being slowly added by the mixture control. The Mixture is open fully for a few seconds first to allow any built up vapor pressure to purge out, then its closed; this gives liquid fuel a clear path down the lines. The mixture is kept closed until the cranking so that you have complete control of how much fuel to add to get the engine lit off - which likely will be different every time - this technique also significantly reduces the changes that you will flood the engine. There is NO boost pump used because the mechanical pump should be able to provide fuel at a rate that keeps excess fuel low, unless you have a hot soak condition... You will know if you have a hot soak event because the above technique will not work after two tries. Summary: There is liquid fuel available at the fuel pump but there is vapor in the injector lines. The vapor does not ignite easily and it blocks liquid fuel from flowing past the vapor area (vapor-lock) causing a hard to start condition. You have to relieve the pressure of the vapor and then slowly add liquid fuel to the lines and the injectors so that the mixture of fuel and air is correct for ignition in a hot cylinder.

Long Soak:

  1. Throttle Closed
  2. Mixture Full Rich for about 5 seconds then Idle Cut Off (Be very sure it is fully at idle cut off)
  3. Boost pump on (or on low if two speed) for 30-60 seconds. (yes, a full half a minute to a minute)
  4. Boost pump OFF
  5. Throttle cracked to the point where it would need to be to have about 1000-1200RPM if the engine were running
  6. Begin cranking the engine, wait a 2 seconds then slowly (over 5-10 seconds) move the mixture toward rich. Don’t exceed recommended cranking time.
  7. As soon as the engine starts, keep the mixture at about that point, adjust it and throttle for smooth operation. You should ALWAYS run the engine as lean as possible on the ground.

Rationale: The Mixture is open fully for a few seconds first to allow any built up vapor pressure to purge out, then its closed; this gives liquid fuel a clear path down the lines. You want to be very sure that the Mixture is fully at idle cutoff because we do not want any fuel to get past the metering circuit. Running the boost pump with the mixture closed will pressurize the fuel circuit and circulate some liquid fuel which will help cool things down and reduce additional vaporization. Excess vapor will be expelled through a vent port and the mechanical pump and the lines leading to the metering circuit will become fresh with cooler fuel. This takes time, and you have no worry of flooding the engine because the mixture is at idle cutoff. You finish by following the same procedure as a short soak. Summary: The mechanical fuel pump has to have the engine cranking to do its job. It would take too long and be too hard on the starter to use this pump to purge vapor. (Hence why I’ve seen pilots cranking the engine for absurdly and dangerous amounts of time). The electric pump is parallel* with the mechanical pump and has the ability to run quickly and more efficiently to do the job of purging the vapor, but its not an easy job - it takes a lot of time therefore you need to run the pump for at least 30-60 seconds. Once the vapor is purged, liquid fuel can flow past the metering circuit, and now you have the situation of a short soak to deal with, see above.

So why not just use a Hot Soak procedure every time?

You could, but after you understand why the damn thing won’t start and you think about it, you can use the technique that works the best.  If I land the plane, shut down hot, and then go for a restart in less than 10 minutes or so (all thing considered equal), I know that my problem is not at the pump, its at the top lines, so why waste time.

Disclaimer: Follow your POH unless you fully understand the pro’s and con’s of using other methods and other’s advice.

DVA

*As Don Kaye correctly pointed out, the electric fuel pump (on the M20M) is physically in “series" with the mechanical pump as shown on the schematic. The point is that neither of the two are dependent on one another and both can participate alone or together, in parallel, to provide fuel flow and pressure. 


1 person likes this


11 Comments


Do you have any CHT numbers to determine a hot start vs a cold start. A hot engine at Goose Bay in the winter will cool quicker than one in Miami.

José

Share this comment


Link to comment
30 minutes ago, Piloto said:

Do you have any CHT numbers to determine a hot start vs a cold start. A hot engine at Goose Bay in the winter will cool quicker than one in Miami.

José

Hi Jośe, always an honor to interact with you. 

The short answer is no. I think this is more about how hot it is outside after you shut it down verses how hot the engine is at any given time. Except of course for the short soak, which I think will occur regardless of the outside temp for the most part.

You're very experienced, what have you found?

Share this comment


Link to comment

That is a good description of hot starting a Lyc, Fuel injected engine, I use your first method all the time, and add boost pump after running if needed to clear on a "long soaked" condition. Continentals, a different animal with a different injection system

Share this comment


Link to comment

I never had any problem with hot starts but most have been done just after refueling in warm weather. Never looked at the CHT before start. But there are instances in other countries were refueling/payment/clearance can take up to an hour after engine shutdown. Is this a hot or cold start situation? I think outside temp is a factor to consider, that is why I think knowing a CHT threshold for hot vs cold would help. When in doubt I do a hot start first, if no luck I do the cold start. It has always worked for me. Just a wild guess; anything below 150F CHT would be a cold start. What do you think?

Edited by Piloto

Share this comment


Link to comment

Thermocouple in the fuel spider would be an interesting sensor...  (Low cost and simple)

There is a temperature range that would cause some fuel to boil.  The hotter, the more likely the need for hot start...

a thermocouple mounted where the fuel return to the tank is, would indicated if the 30 seconds of running the pump with mixture out is doing anything useful.

Thinking out loud...

Best regards,

-a-

Share this comment


Link to comment

The engine's fuel has boiled from the lines and into the intake ports, thereby6 flooding the engine on a hot start. But your method has us priming it even more, flooding it further. 

The Dukes fuel pump is closing in on nearly two grand to overhaul, and running it for 60 seconds as a standard procedure will finish it off pretty quick. 

Also, you can run the boost pump as long as you want, but it wont purge the fuel injector lines. That vent line should never vent fuel, or vapor. its only there to dump fuel overboard in case of a torn diaphragm in the fuel pump.  But cranking it with the boost pump on will quickly fill the mechanical fuel pump and purge it pretty fast.   It pretty much starts right away, no additional priming, and no sitting there with the fuel pump running. 

Edited by jetdriven
2 people like this

Share this comment


Link to comment

Thanks for the comments, jetdriven!

First I'd like to point out that this is not necessarily "my method", it is simple one that has worked for many people over many years and it makes sense in many cases. For some it may not work or it may not make any sense.  Then, just for an abundance of caution, I added: Disclaimer: Follow your POH unless you fully understand the pro's and con's of using other methods and other's advice.

22 hours ago, jetdriven said:

The Dukes fuel pump is closing in on nearly two grand to overhaul, and running it for 60 seconds as a standard procedure will finish it off pretty quick.

 Are you referring to a boost pump on a Beech Duke? Or a Duke’s Electric fuel pump?

22 hours ago, jetdriven said:

The engine's fuel has boiled from the lines and into the intake ports, thereby6 flooding the engine on a hot start.

Heat is culprit here, and there is likely no full “boiling” occurring in a short heat soak. The various hydrocarbons in gasoline (AVGas included) “boil" at different temperatures with some very low around 90dF and others very high at over 280dF. Fuel does not have to boil (fully or partially) to evaporate into a vapor. In a vapor lock, the fuel is simply turning from a liquid to vapor - how it gets that way is largely irrelevant. Could it have a boiling component? - Sure and during a long heat soak it is very possible and that would exacerbate the problem, but boiling is not necessary to create a vapor lock.

The fuel in the feeder lines to the injectors, and any fuel in the intake manifold or ports are all vaporizing due to the high stagnate heat, the liquid is changing to a gas, the gas is less dense and the molecules are not of the correct atomization size to mix with the air. If you try to fire the engine in this state, you have a lean condition. 

Sidebar: A fuel’s measure of Volatility is important here. Petro-chemists adjust a fuels volatility for different applications; in piston aircraft the volatility index or rating needs to be high enough to allow for a relatively low pressure feed system, it needs to be correct enough for both carburetor and fuel injected engines, and it must be correct enough to work well from SL to about 25K in a wide temperature range. That’s quite a requirements spec and AVGas does this fairly well. The higher the volatility the easier it is for the liquid to turn into a vapor or to atomize well with air which is good for our very unsophisticated and crudely simple airplane motors that operate in myriad conditions. High volatility is also a Bad Thing because it allows liquid fuel to vaporize very quickly at rest, in higher temps, at low pressures, which causes hard starting.

22 hours ago, jetdriven said:

But your method has us priming it even more, flooding it further. 

If you think about, it’s just the opposite. The method described for starting a short heat soaked engine is to first introduce air, and then to slowly add fuel with the mixture control.  For a long heat soak, the mixture in cutoff while the rest of the system is pressurized, so no fuel is flowing to the cylinders, there is no chance of flooding anything.

22 hours ago, jetdriven said:

Also, you can run the boost pump as long as you want, but it wont purge the fuel injector lines. 

Right. The Long soak procedure where that is recommended is not intended to purge the injector lines, the mixture is at idle cut-off. The boost pump is run to purge the lower part of the system and the pumps themselves. You purge the injector lines during the short soak procedure. Thanks for pointing that out, sorry if I wasn’t clear on that.

22 hours ago, jetdriven said:

That vent line should never vent fuel, or vapor. its only there to dump fuel overboard in case of a torn diaphragm in the fuel pump.  But cranking it with the boost pump on will quickly fill the mechanical fuel pump and purge it pretty fast.   It pretty much starts right away, no additional priming, and no sitting there with the fuel pump running. 

The vents are there for both reasons. A un-primed fluid pump (like a fuel pump) will not self-start, it needs to be primed - ask any firefighter who operates the pumper-tanker the theory is exactly the same. If the liquid fuel has (maybe as you surmise) boiled away from the intense heat trapped in a closed cowling, the pump(s) need to be refilled so they can push fuel again. The reason for using the electric pump is that it requires no engine cranking, and operates at a higher initial pressure in most cases. It’s far easier to get that fast spinning electric pump to get a little fuel flowing that it is to crank and crank and crank on the starter in the hopes of getting the mechanical pump to prime.

Again, this is just one way to overcome a problem that has worked for many. Your milage will vary. But I have seen a number of very pleasant ‘a-ah!’ moments when these procedures we used where others may have failed.

Thanks for the dialog!
DVA

1 person likes this

Share this comment


Link to comment
11 hours ago, DVA said:

Thanks for the comments, jetdriven!

First I'd like to point out that this is not necessarily "my method", it is simple one that has worked for many people over many years and it makes sense in many cases. For some it may not work or it may not make any sense.  Then, just for an abundance of caution, I added: Disclaimer: Follow your POH unless you fully understand the pro's and con's of using other methods and other's advice.

 Are you referring to a boost pump on a Beech Duke? Or a Duke’s Electric fuel pump?

the Dukes elecric pump installed on Mooneys is ridiculously epensive, around 1400$ plus freight plus install to overhaul. It has plastic vanes and a carbon shaft seal, we ruined a fresh one in a single hour.  Seems they dont like to run continuously. Anyways, the less they run the better.

Heat is culprit here, and there is likely no full “boiling” occurring in a short heat soak. The various hydrocarbons in gasoline (AVGas included) “boil" at different temperatures with some very low around 90dF and others very high at over 280dF. Fuel does not have to boil (fully or partially) to evaporate into a vapor. In a vapor lock, the fuel is simply turning from a liquid to vapor - how it gets that way is largely irrelevant. Could it have a boiling component? - Sure and during a long heat soak it is very possible and that would exacerbate the problem, but boiling is not necessary to create a vapor lock.

The fuel in the feeder lines to the injectors, and any fuel in the intake manifold or ports are all vaporizing due to the high stagnate heat, the liquid is changing to a gas, the gas is less dense and the molecules are not of the correct atomization size to mix with the air. If you try to fire the engine in this state, you have a lean condition. 

I dont think so. the engine is flooded on a hot start becuase the boiling fuel int he stainless injecotr lines push liquid fuel through the injector and flood the cylinder.  Hence the "no touch 1000 rpm" starting method. simply crank with no mixture after 10-20 blades it fires up. It wont often stay running due to vapor lock, but a quick touch of boost pump fixes that.

Sidebar: A fuel’s measure of Volatility is important here. Petro-chemists adjust a fuels volatility for different applications; in piston aircraft the volatility index or rating needs to be high enough to allow for a relatively low pressure feed system, it needs to be correct enough for both carburetor and fuel injected engines, and it must be correct enough to work well from SL to about 25K in a wide temperature range. That’s quite a requirements spec and AVGas does this fairly well. The higher the volatility the easier it is for the liquid to turn into a vapor or to atomize well with air which is good for our very unsophisticated and crudely simple airplane motors that operate in myriad conditions. High volatility is also a Bad Thing because it allows liquid fuel to vaporize very quickly at rest, in higher temps, at low pressures, which causes hard starting.

Preach on brother. Amen.

If you think about, it’s just the opposite. The method described for starting a short heat soaked engine is to first introduce air, and then to slowly add fuel with the mixture control.  For a long heat soak, the mixture in cutoff while the rest of the system is pressurized, so no fuel is flowing to the cylinders, there is no chance of flooding anything.

Its already flooded. However, it goes from flooded to lean cutoff, due to the whole fuel system filled with boiled vapor.

Right. The Long soak procedure where that is recommended is not intended to purge the injector lines, the mixture is at idle cut-off. The boost pump is run to purge the lower part of the system and the pumps themselves. You purge the injector lines during the short soak procedure. Thanks for pointing that out, sorry if I wasn’t clear on that.

The vents are there for both reasons. A un-primed fluid pump (like a fuel pump) will not self-start, it needs to be primed - ask any firefighter who operates the pumper-tanker the theory is exactly the same. If the liquid fuel has (maybe as you surmise) boiled away from the intense heat trapped in a closed cowling, the pump(s) need to be refilled so they can push fuel again. The reason for using the electric pump is that it requires no engine cranking, and operates at a higher initial pressure in most cases. It’s far easier to get that fast spinning electric pump to get a little fuel flowing that it is to crank and crank and crank on the starter in the hopes of getting the mechanical pump to prime.

The electric boost pump will prime a mechanical pump very quickly. Its very useful for a hot start where it will hit but not stay running. But I'd state that any fuel that comes from the electric pump drain or mechanical pump drain renders it unairworthy. They are a telltale, not really like a sniffle valve.

Again, this is just one way to overcome a problem that has worked for many. Your milage will vary. But I have seen a number of very pleasant ‘a-ah!’ moments when these procedures we used where others may have failed.

Thanks for the dialog!
DVA

A pleasure.   Byron

A

Share this comment


Link to comment

I just replaced my Weldon electric pump on my M20J. It came out to be 1085$ plus $400 core at Aircraft Spruce. The old pump was the original one 1982. What is the TBO on this pumps? After all you only use it for a few seconds for every hour you fly. Assuming you pump for 5 seconds for one hour flights that turn out to be 2.7 hours pump time for 2,000hrs engine time. And the pump time could be much less for longer flights. The electric fuel pump on some cars and jet planes is next to the fuel tank and runs continuously.

Edited by Piloto

Share this comment


Link to comment
9 hours ago, jetdriven said:

DVA>The fuel in the feeder lines to the injectors, and any fuel in the intake manifold or ports are all vaporizing due to the high stagnate heat, the liquid is changing to a gas, the gas is less dense and the molecules are not of the correct atomization size to mix with the air. If you try to fire the engine in this state, you have a lean condition. 

I dont think so. the engine is flooded on a hot start becuase the boiling fuel int he stainless injecotr lines push liquid fuel through the injector and flood the cylinder. 

Byron, I’ll offer that the increase in pressure in the fuel circuit, created by the vaporizing will take the path of least resistance to relieve itself. And that could also be away from the intake ports, and back into the fuel circuit. If it’s hot enough to vaporize the fuel in the injector lines, it is certainly hot enough to continue to vaporize that fuel in the cylinder, resulting in no liquid - therefore no flooding. That said, my experience is not in aircraft engines, but in old school automotive where the above is un-argued. I am looking for a hint as to why there is such a difference in an aircraft ICE to help me see the light and have a wonderful ‘a-ha!’ moment. 

Based on that experience, I maintain that the issue of a hard start is combination of not enough fuel (lean), in the right composition (not atomized and suspended in the air), not at a constant rate (based on vapor locking prevent even “liquid" fuel flow) to allow for a normal proper and quick start.

We might be able to do an experiment to see what really happens in the cylinder?  Would it be logical to say that if the theory of a rich flooded condition occurs, and liquid fuel floods the cylinder(s) as you wrote above, that you should be able to detect that via borescope inspection after a period of time post a hot shutdown?

Dave

Share this comment


Link to comment

 I have never experienced "vapor lock" in my Lycoming IO360 (Bendix RSA fuel system).  I have experienced fuel boiling in the lines after shut down. This symply "pre-primes" the engine for the next start. There is no need to do anything but add air and spark; once the engine fires off, slowly introduce more fuel.

Form what you've written below, I am guessing your advising folks on how to "hot start" a Continental engine.  The fuel systems differ significantly from the Bendix systems found on Lycoming engines. Running the boost pump with the mixture set at idle cut off is useless if trying to start a Lyc with a Bendix fuel system.  Some of the advice you're giving would make starting a hot Lycoming more challenging.

Long Soak:

  1. Throttle Closed
  2. Mixture Full Rich for about 5 seconds then Idle Cut Off (Be very sure it is fully at idle cut off)
  3. Boost pump on (or on low if two speed) for 30-60 seconds. (yes, a full half a minute to a minute)
  4. Boost pump OFF
  5. Throttle cracked to the point where it would need to be to have about 1000-1200RPM if the engine were running
  6. Begin cranking the engine, wait a 2 seconds then slowly (over 5-10 seconds) move the mixture toward rich. Don’t exceed recommended cranking time.
  7. As soon as the engine starts, keep the mixture at about that point, adjust it and throttle for smooth operation. You should ALWAYS run the engine as lean as possible on the ground.

2 and 3 above would do nothing and I mean nothing for a Bendix fuel system. On a Continental it would push fresh fuel to the servo and the hot fuel would go back in the tank via the fuel return. 

Share this comment


Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now