Menu Sign In Contact FAQ
Banner
Welcome to our forums

Fuel servo icing

Well, the 10K delta T from the OAT to the servo inlet is a fact. How exactly it is made up (compression, conduction, radiation) doesn’t matter. You can get a feel for the possible contributions of these factors by looking at the graph and seeing how the delta T varies with IAS. The climb would have been about 95kt IAS, the descent about 150kt IAS.

Administrator
Shoreham EGKA, United Kingdom

Peter wrote:

Well, the 10K delta T from the OAT to the servo inlet is a fact

Technically you have not measured the delta T. You have only measured the temperature at two places with two unrelated thermometers. The industrial way of doing this is a matched pair of calibrated RTDs after IEC or directly with a pair (or more) thermocouples. Besides, measuring OAT has to be done ensuring no water/ice or condensation on the probe, or you will on average measure something closer to the wet bulb temperature rather than the dry bulb temperature, which can explain the 10 degrees and more. And of course the heat from the engine will affect things.

Your readings are facts, but have you actually measured what you think you have measured is what I am wondering about.

The elephant is the circulation
ENVA ENOP ENMO, Norway

The industrial way of doing this is a matched pair of calibrated RTDs after IEC or directly with a pair (or more) thermocouples

The probes were PT100 and these are normally matched to 0.1C. You can see their matching at the start of the graph as being within 1C

but in reality they are far closer. I know which data logger was used for this test.

The servo inlet sensor was carefully mounted to measure the air and not the duct.

The data confirms my experience that with an OAT of -15C the alternate air door will eventually be found to be frozen solid, and it will be thus with the airframe completely free of any visible ice. I think most pilots will find this staggering but it’s true, and it is going to make you think if your alternate air door is spring loaded with no manual control!

The delta T does vary with the airspeed but if you look at the curves (I suspect nobody has – it’s a bit like sending somebody a 100 line email ) you see that it is smaller during the descent so any compression heating is only a part of it, otherwise it would be the other way round.

I still find it staggering that you get a 10C rise in that short duct. The air velocity must be substantial, with the engine sucking like a €10 ********* and given that the air filter does not get blocked with water even in the most powerful rain imaginable.

measuring OAT has to be done ensuring no water/ice or condensation on the probe, or you will on average measure something closer to the wet bulb temperature rather than the dry bulb temperature

I also have handwritten notes from the flight which show the real OAT probe alongside the logged OAT, and there is correspondence (a 3C or so difference) which happens to have been confirmed on a test done several years ago for another purpose. The whole logged flight was in VMC so there is no reason to think there was a lot of water anywhere.

Administrator
Shoreham EGKA, United Kingdom

Peter wrote:

The whole logged flight was in VMC so there is no reason to think there was a lot of water anywhere.

And yet the alternate air door is frozen solid?

Look at the custom made RV’s made for speed. They use a special cold induction sum with a forward facing intake to match directly to a short and straight induction tube under the cowling. It’s to get the air into the engine as cold as possible and with a little resistance as possible. This wouldn’t have evolved if the “standard” setup worked well, ie the standard setup heats up the air by conduction. The 10 degree difference you see (if correct) could therefore be used for more HP, and additional HP with the cold air sump. The “compression heating” is the same (2-3 degrees), but could actually (almost) be eliminated, at one particular speed at least, by careful sizing and shaping of the air intake.

The elephant is the circulation
ENVA ENOP ENMO, Norway

And yet the alternate air door is frozen solid?

If you fly in IMC, yes.

That data was collected in VMC, because collecting it in IMC is vulnerable to various hazards, plus you can get measurement errors due to ice formation, sublimation, liquid water condensation/evaporation and the usual latent heat related effects.

I am not for a moment doubting that somebody may have been through this exercise before, but they sure as hell don’t seem to have published their data.

Incidentally, while the TB20 POH asks for using alternate air in icing conditions, it contains this

which IMHO is very bad advice, unless “icing area” = IMC.

Administrator
Shoreham EGKA, United Kingdom

Peter wrote:

but they sure as hell don’t seem to have published their data.

If you search around the forum at VansAirforce there may be some stuff (entangled within lengthy threads), but I doubt many (if any) has done actual measurements other than speed tests. Top speed in dry desert air is not very relevant to European conditions, but the solutions are. What is done is to increase the thermodynamic efficiency of the engine, and that can be used for both added HP and better fuel economy. This is the first such measurements I have ever seen.

The elephant is the circulation
ENVA ENOP ENMO, Norway

I just came across this video. They talk about this subject around minute 9



Feel free to remove if already posted

Antonio
LESB, Spain

Yeah – he’s got it right.

I got a lot of “trouble” here at the time from some “experts” but they were wrong Later it was proven with an instrumented TB20 and a test flight.

The only thing this guy gets wrong is the comment “freezing rain”. Any air, with any moisture, below 0C at the fuel servo entry will ice up those tubes. And on the TB20 you get -5C there with an OAT of -15C (or so). But it doesn’t always ice up – just like normal structural ice doesn’t always accumulate.

Don Rivera is the guy who rebuilds my RSA fuel servos! For mine he does an 8130 also.

Great video – thanks for posting it!

Administrator
Shoreham EGKA, United Kingdom

Yes the guy knows these servos in and out. Other than the icing issue, I like these servoed fuel injection systems on Lycos better than the Continental mechanically metered flow systems on NA engines. The former are up to a limit independent in their function from incoming fuel pressure, whereas it does affect fuel flow on the latter. Of course the latter are more robust and immune to icing and iallow supplemental fuel flow through use of the boost pump, which comes in handy at times in the turbo variants.

Last Edited by Antonio at 23 Nov 22:57
Antonio
LESB, Spain
79 Posts
Sign in to add your message

Back to Top