Hot start with a PT6A-67D

[mnttech]

Hi,
While working on a Beech 1900D simulator, the NAA said that the hot start malfunction should allow the engine to actually start. Our model ramps the ITT very rapidly to 1000°C and stops the N1 at about 30%. While I have looked at the postings here about not getting a hot start, there has been very little about what does happen. Anyone out there who would like to share what happens when a PT6 does hot start?

Thanks!

[V1... Ooops]

MNT Tech:

If you are associated with FlightSafety in any way, contact the Toronto Learning Center and ask them to provide you with the code for the hot start malfunction from the Twin Otter simulator. Although it is for a PT6A-27 engine, it will be suitable for the -67. That code has been carefully optimized to provide a realistic hot start.

The most probable cause of a hot start is late ignition. Thus, if you are optimizing your code, light-up should be later than normal. Another probable cause is NG stabilizing at a much lower than normal value (e.g. 12% or similar), then the pilot introducing fuel.

My own belief is that the gas generator should continue to accelerate normally during the hot start.

[mnttech]

No to the Flight Safety, but thanks very much. Have you ever actually had a hot start?

[Cardinal]

The Aims or Satco sim?

I concur with your NAA, the engine should start. I've had a few -67D starts go poorly. All the hot starts I've witnessed there was increasing N1, unfortunately ITT was increasing much faster than normal due to a weakening battery or tailwind. The engine would eventually have reached a sulf-sustaining N1, but had the start not been aborted limitations would have been exceeded and permanent damage may have been done. With the perfect tailwind I even had a T-handle illuminate once, but the start was just within limitations.

Hung starts I've had also. Cold day, weak battery, N1 accelerates to about 35% after ignition then stagnates, ITT about 500, starter still engaged, and little to no progress. The 20 sec. limitation on the starter doesnt apply after light off so you can just watch and wait. If N1 is increasing at all the engine will eventually spool up rapidly above 45% or so. If after about a minute of stagnation there isn't any progress the start is probably hopeless.

Your current model of a hot + hung start is a very bad day indeed. In nearly 4000 hours on the airplane I never ran into such a circumstance. The only scenario I could imagine would be an extremely weak batt, so weak that 12% N1 couldn't be attained, then a ham fisted pilot introduces fuel at something like 8% and just leaves it there. That might? produce such extreme temps and stagnant N1. (Now the true arctic warrior would modulate the condition lever to moderate ITT while nursing N1 higher, but that's a discussion for another day.)

[aerobat77]

i can only speak for the pt6a-41 on the cheyenne III i ride but i must say start malfunctions on start are rare.

one start get hotter than the other , but generally- keep cool. spin the core until no further increase in ng is possible, kick in and look. the ITT may shoot up, the compressor increase slow, but it mostly stabilises.

when the engine does not reach itt limits i do not abort and it mostly start to run.

a hung start is very rare, but may happen. also here, do not abort to quick, look what the engine does.

i had also a wet start due to a faulty ignitor box- ok, here you can stop and call for a mechanican.

at 1000 deg i keep on going- limit is 1090 :-)

[bbrunton]

Hot Start? I have never seen a start get to the start temperature limit if all is normal.

If the battery is weak, and the N1 does not reach 12% there is no reason to introduce fuel at all. If you try there will probably be no fuel flow as there is not enough N1 to produce enough fuel pressure to open the minimum pressurizing valve.

Acessory gearbox flooding - too much oil in the acessory gearbox, dragging on the gears there can be a problem. Probably in really cold conditions. That will drag down the N1 and slow the acceleration. I have never seen it but I have heard of it.

A tailwind is not much problem. It usually just causes the exhaust to be injested into the intake and cause the bleed air to stink up the cabin. Bleed air should be off during start anyway. No use taking air away from the engine during start. The cabin sure does not need it then.

If the start hangs at about 30% N1 that usually means the secondary fuel nozzels did not receive any fuel due to a stuck transfer valve. The primaries usually do not supply enough fuel to get that hot by themselves. That increase in acceleration that happens at about 35% is the secondaries opening up.

A milking start - Turning the fuel on and off to control temperature is a NO NO! If the engine and battery are in good shape, I have never seen a start go very much past the normal running temperature limit during start. And that is only on a hot day with a tailwind! 99% of starts don't exceed the normal running red line, and the temperature is moving upward very slowly and about to peak if it does exceed the normal red line.

If you have a weak battery, a tailwind, bleed valves open, you introduce fuel at exactly 12%, and you have any engine problems like stuck bleed valve, bad guide vanes, or other engine damage, then the start could get close to the start limit.

In closing.... If during start the ITT is passing the normal running red line and is not slowing down - ABORT THE START!!!!!! Find out what is wrong.. It should not do that..

I have made thousands of PT6 starts in just about every dash number from -6 to -67 and I have never had a problem... Well there was one problem... A -67A during a windmilling air start that decided to accelerate and then the N1 suddenly dropped to 0%. It seems the N1 Turbine was squeezed by the shroud because the shroud did not heat evenly and distorted!! One hot section inspection, all new blades on the T wheel, and $40,000 later, all was well again. It was not my fault. I did everything correctly.

[SNS3Guppy]

The most probable cause of a hot start is late ignition.

The most probable cause of a hot start is low battery or a weak starter. Too low an operating RPM at the time fuel and ignition are introduced tend to lead to a start going hot. The primary reason is that the engine isn't turning at a high enough RPM, and has insufficient fuel flow.

While minimum RPM's are provided for the introduction of fuel during the start sequence (movement of the condition or fuel lever, depending on your installation), one is always smarter to wait until the engine has reached it's maximum motoring RPM during the start sequence before introducing fuel.

Will an engine actually start during a hot start sequence? You should never find out, and it's irrelevant. If the engine has gone hot, it needs to be shut down. If it's approaching start limits, then fuel should be cut off and the engine motored...then started normally if possible, to stabilize internal temperatures. If the engine exceeded limits, it should be cut off, motored to cool, and shut down....and not run again until inspected.

What you should not do is let the start continue if the engine has gone hot. This is a critical condition, and as far as you should be concerned, marks the end of the life of that engine during the time you'll be operating it. Don't let the engine approach a takeoff limiting temperature setting and do nothing. When you see temperatures accelerating rapidly, cut off that engine immediately. If you see temperatures approaching limits, cut off that engine immediately. Don't let it exceed temp limits in the first place.

Causes for hot starts can be several; bleed failures, certain reference line failures, fuel control problems, case drain failures, poor starter, early introduction of fuel, low battery, damaged compressor, inlet blockage, etc.

The damage that can be done from an engine going hot won't be readily apparent, but can be excessive,and can lead to a catastrophic failure.

Not all engines, or even all dash numbers of engines use the same thermocouple placement. While an engine may appear okay, the actual temperatures inside may be far hotter in other areas of the engine where temperature monitoring is not taking place, and where temperatures normally don't become a problem. This is the case with a hung start, when the engine can be cooking farther upstream from the thermocouples. The engine may not appear to be getting that warm, but it may in fact be burning up upstream from the thermocouples, with in sufficient heat getting to the thermocouples early enough to warn you. Your only warning may be acceleration stagnation, and there are a number of reasons that the engine can reach a hung start condition.

A milking start - Turning the fuel on and off to control temperature is a NO NO!

Depends on the engine. I operated a -45R for some time in which it was absolutely necessary. It's a common requirement in some engines, particularly some C250 engines at higher density altitudes. Some engines do fair better, and have a longer life, or simply require an operator-modulated start. I've never had to do it on a -67 that I can recall, but others, yes.

Have you ever actually had a hot start?

Absolutely. And hung starts. Pay close attention to the engine instrumentation during the start. What you miss or don't address during the start could very easily come back to bite you, or someone else later, during takeoff or in flight.

[V1... Ooops]

SNS3Guppy:

What does your pontification above have to do with the original poster's question?

[V1... Ooops]

MNT Tech:

Concerning development of simulation for hot starts – I have done this before on a FSI sim, and here are some thoughts that might assist you:

First, you should get together with your instructor team and determine exactly what it is that the instructors want to accomplish with this particular malfunction. Do they want the malfunction to have characteristics such that the pilot should abandon the start attempt (cut off the fuel), then continue to crank the engine? Or, do they want to present a starting scenario in which the start temperatures are higher (for cause) than what is normally observed, but don’t exceed the published limits? In the latter case, it would be appropriate for the student to continue the start, most especially if the lesson plan setup puts the aircraft at a remote location where no ground power or maintenance services are available.

I’m not familiar with your aircraft or engine model, but I am very familiar with the PT6A-27, so, I’ll use limitations for the -27 in this discussion – you can easily apply the concepts to your aircraft by modifying the limitations to match your aircraft’s AFM and engine MX manual.

There is an enormous amount of confusion and misunderstanding amongst pilots of PT6 powered aircraft about what exactly constitutes a “Hot Start”. So, let’s first define a Hot Start as being a start sequence where one of the published limitations that are applicable during the starting phase of engine operations are exceeded. This definition does not include starts wherein the starting temperatures are much higher than normal (usually for cause), but do not exceed the published limitations that are applicable to starting conditions.

For example, in the case of the -27 engine, the starting limitations permit temperatures of up to 1090° provided that:

1) the temperature drops below 980° within 2 seconds, and;
2) the temperature continues to drop to below 925° within the following 8 seconds.

The illustration below – taken from the engine maintenance manual - depicts these limits.

http://i979.photobucket.com/albums/ae275/Paneuropean/StartingLimits.jpg

As you can see, temperatures of up to 925° for the entire duration of the start are acceptable. Acceptable does not mean desirable, nor does it mean common, ‘acceptable’ simply means that the starting temperature limitations have not been exceeded.

Only a minority of pilots are aware of these starting limitations, because unless the aircraft is fitted with an electronic display system that automatically adapts the coloured green, amber, and red instrument markings to suit the phase of operation of the engine (starting, idle, take-off, etc.), the only limitations that the pilot sees are those applicable to take-off, cruise, and (perhaps) MCP. The picture below shows the T5 gauge on an aircraft fitted with a -27 engine. There is no indication of the starting temperature limits, other than a single red radial mark at the 1090° position. The colour coding on the face of the gauge corresponds to the temperature limitations that are applicable in conditions other than starting. The top of the green arc is the limitation for climb and cruise, and the top of the amber arc is the limitation for take-off.

http://i979.photobucket.com/albums/ae275/Paneuropean/T5.jpg

Normally, if the battery is in reasonable condition or if an external power cart is used, the highest temperature encountered during any portion of the start will not result in a temperature higher than the green arc depicted on the gauge. This is coincidental. As a result, the pilots have a habit of thinking that “green is good” so far as start temperatures are concerned, and if the peak start temperature results in a gauge reading in the amber or red region, that must indicate a caution or warning condition applicable to the start. That is not correct.

For sake of this discussion, let’s describe any start with a peak temperature greater than the green arc on the T5 gauge as being a “warm” start, with “warm” meaning higher than normal, but not exceeding starting limitations.

As you are probably aware, there is a direct and (almost) linear correlation between peak starting temperature and the stabilized NG achieved prior to introduction of fuel. So, before you even begin to model the malfunction, make sure that the engine library is written such that it models this behaviour. A stabilized NG of 22% prior to fuel introduction (correlating with a good quality external power cart) should give a peak start temperature of (for example) 625°, a stabilized NG of 18% (correlating with a fully charged battery) should give a peak start temperature of perhaps 675°, and a stabilized NG of 12% (correlating with a seriously discharged battery) should give a peak start temperature of about 900° to 1,100°. Don’t forget to ensure that your battery library reflects the current drain that the first start takes out of the battery – the second engine should achieve a stabilized NG of about 1% less than the first engine if both engines are started sequentially from the battery alone. Thus, using the battery only for starting, the peak starting temperature of the second engine should be slightly higher than the first.

There can be many causes for a warm start. The most common is lower than average NG prior to introduction of fuel. Let’s presume that you write the code such that achieving a stabilized NG of 14% prior to fuel introduction (correlating to a battery in poor condition, or a partially discharged battery) gives a peak starting temperature of 800°. That’s easily within starting limitations, but certainly much higher than pilots will see on a daily basis. This is why (and where) you need to talk to your instructors and find out what behaviour they want to see from the pilot. If your instructors foresee a scenario where the student is briefed that the aircraft is at an out station with no maintenance capability, it’s late at night, and an ice storm is forecast to come through in 2 hours… and, oh, by the way, the battery is quite discharged as a result of (whatever – talking on the radio, passenger getting sick before engine start and needing to be offloaded, etc.), then the logical pilot action would be to anticipate that the stabilized NG prior to fuel introduction will be much lower than normal, and consequentially, the start will be much warmer than normal, peaking at 800° or so. That’s not a problem, and in the case of the -27 engine, doesn’t even need to be reported (the AFM and MX manual suggest that starting temperatures greater than 850° be investigated for cause, although in this scenario, the cause is well understood before the start is even commenced). This type of scenario presents a ‘warm start’.

On the other hand, let’s say your instructors want to present a scenario in which the starting limitations are well and truly exceeded. This might be caused by a fault in the ignition system that allows 5 to 8 seconds worth of fuel to pool in the combustion chamber prior to ignition. In such a case, there will be a huge spike in T5 following light-up, and it would be realistic to expect that the T5 temperature indication might peak above the 1090° limit, or, it might rise up to 980 and then hang there for longer than the 2 seconds allowed. This type of scenario would present a true “Hot Start”, that being defined as starting temperature limitations being exceeded. The expected student response would be to shut off the fuel and continue to crank the engine (assuming that is what the AFM for your aircraft states should be done).

So, to summarize, before you can write the code to deliver a realistic “Hot Start”, you first have to make sure that your normal code varies peak starting temperature in relation to what stabilized NG is achieved prior to fuel introduction, and you then have to get your instructors to tell you exactly what the learning objective of their “Hot Start” scenario is. Once you have the engine giving good simulation of starting temperatures relative to air density and stabilized NG prior to fuel introduction, you can create a set of conditions that will give a true “Hot Start”. But… don’t forget that simulation is theatre, and good theatre requires realistic presentation. It would not be realistic to have the engine spool up to normal NG, and light up in the normal time frame after fuel is introduced, then hammer the student with a starting temperature higher than limitations. You need to build in a credible reason for the hot start. That could be delayed ignition. It could also be a result of fuel pooled in the bottom of the engine following a prior aborted start attempt… but doing it that way requires the instructors to set the stage in their briefing (by explaining that the prior start attempt did not result in light-up and was abandoned). Personally, I think it’s easier (and more realistic) to use delayed ignition.

You could also set a low NG threshold that causes the peak starting temperature to exceed limitations. For the -27 engine, pilots are advised to not introduce fuel unless they achieve 12% stabilized NG. So, you could write your code such that the engine only stabilizes at 11% NG, and if the student then introduces fuel, the starting temperatures then exceed the published limitations.

Have fun with your project…

[SNS3Guppy]

What does your pontification above have to do with the original poster's question?

What does answering the poster's question about hot starts have to do with answering the poster's question about hot starts?

You asked this question?

[aerobat77]

i experienced that the pt6a vary significant with the starting temperature due to conditions . a very quick turnaround at the airport on a hot summer day may give a hotter start, an airport in the alps with high elevation also, its also true that battery condition affects the temps very much.

i saw starts on the -41 where ITT went to nearly 1000 but stabilized quickly and the engine made a normal start.

there are also starts where the temps do not exeed the normal operating redline.

when the engine of mntech jumps everytime to 1000 there is defititly something wrong, when just occasionally we may talk here a non existing problem since it may happen on accasions - when it peaks at about 1000, drops again a little and stays just below 1000 i do not abort but keep on starting.

[mnttech]

Guppy,
Great insight and I appreciate your time for posting. I liked “What you miss or don't address during the start could very easily come back to bite you, or someone else later, during takeoff or in flight.”

V1…Oops
Thanks very much for a great post, and I liked the graph and picture. The discussion around what my instructors want is wonderful if the simulator was used by just one customer, but it is used by several. So I can very easily run into customer A wants this, and customer B wants that, and the NAA wants both, but leans towards a third option.

This topic was also brought up during the last FSEMC in Brighton. There are two lines of discussion around malfunctions.
1. Do we (owner or MFG) have hard data to support any one malfunction, or are we working off of what we “should” see based off of other cases or a lot of engineering research.

2. What are the instructors really trying to teach? A hard and fast procedure out of the QRH or the pilot’s to think?

As everyone has pointed out, there are many variables to a hot start, what one person may say is the hard answer, the next may never have seen.

I have always found these kinds of limits impossible to monitor with the older analog gauges like you showed.
1 the temperature drops below 980° within 2 seconds, and;
2) the temperature continues to drop to below 925° within the following 8 seconds.

The JT3D in our 707’s had this Abnormal:
If EGT entered the 450° - 530°C range for 15 seconds (not too hard to do) or less or 530° - 565° range for 5 seconds… bluh bluh bluh.

I have found the errors in the code, my ITTT climbs a lot slower now, the FF does not jump as much, and the engine actually goes above 30%.

My humble thanks to each and everyone of you who took time to post on this.

[SNS3Guppy]

When manufacturers provide tables citing X amount of seconds or X amount of degrees within a particular temperature range, one is playing with fire guessing whether or not the engine has damage. Those tables are really maintenance schedules; they're to determine from a mechanic's point of view, what needs to be done.

Operators may sometimes tell you that the manufacturer allows so long within a maximum temperature range, and that it's okay to go there...five minutes at max temp, for example...what they don't remind you about is that those few minutes could be cumulative over the life of the engine (and even it's subsequent overhauls)...and may have been exhausted long ago.

I once operated an airplane equipped with PT6A-66's, which experienced high temperatures during the start process. My F/O experienced one such event, and I quickly shut it off as the temp climbed rapidly for the limit. Foolishly, I said "here, let me do it," as though my magic touch would make a difference. Not surprisingly, I had to quickly cut it off when the same thing happened again.

The First Officer asked "What shall we do?" I knew exactly what to do. I called the Director of Maintenance. As soon as I told him which tail number I was in, he asked if I was having the temps shoot through the roof on start-up. I said yes, and he said he knew all about it. The airplane had been doing it for three months, he said. I was flabergasted. I asked him what in the name of all things unholy, he thought he was doing continuing to operate that airplane. He told me he had received word from Pratt that so long as the temperature didn't stay there for more than five seconds (well above the start and takeoff limits), he was fine.

I suspected he wasn't telling me everything, having previously been a director of maintenance myself, in an operation flying PT6A's...and I called Pratt directly. The DoM was very wrong; in fact, he was lying to me. I called him back, and he flew into a rage when he learned I'd spoken with the engine manufacturer's technical representatives. He told me I just didn't speak to the right person, and that I should keep asking until I found someone who agreed with him.

Hot starts are very, very serious business. In fact, the most important part of transitioning someone to a turbine engine is teaching them to start the engine. Most operational points after that are a piece of cake. It's the start that does the most damage, most of the time. It's also the one place where a brief moment of inattention can be disasterous, and oh, so costly.