EGT SPREAD

[dlegg]

Im keen on the info regarding Hacman style leaning and thats its working LOP.

I am fortunate enough to have a MGL Voyager Efis with full engine monitoring. This unit has a Gami type engine leaning graph on each cyl. My EGT's are all over the place at full power, but in cruise they do normalise reasonably close, within 20 deg C or so. Using the Hacman valve and the Gami, it is very easy to accurately get 50 C LOP.

 

This is just a screen shot of the simulator but is exactly the same as the instrument. The Hacman is only effective above 5000 or so, but really comes into it's own around 8500 and above. When first starting to use the hacman valve at 8500, at 2900 RPM, revs will increase up to 3150 or so as it leans out, EGT also increasing. Just this by itself is of great benefit. I keep leaning to peak, then LOP by 50 deg C. All the cyls are within 20 deg of each other. Revs will drop down to 2850, CHT's from 150 to 130, EGT's 680 to 700, fuel flow down to 15 LPH.

 

Clearly the engine must be tuned correctly for this to work properly. I have an adjustable prop, ground take off roll RPM is about 2850 and max RPM at 1000 feet AMSL is 3250. This is the best setting for my aircraft, but I do change the prop pitch between summer and winter.

 

I have also taken the air cleaner out of the mixer box, and have a pod filter fitted to the air box inlet, where normaly it would be connected to the naca inlet. My airbox is from Jab USA and is smaller than the normal Jab one, I found that the engine was running too rich with the too small aircleaner. As a result my carby heat air is not filtered.

 

Right now I am very happy with my 3300 and the current configuration will stay as is. I also only run premium 98 mogas and after 340hrs the engine is very clean internally.

 

 

[Oscar]

It won't at no load. Air/fuel volume relates to power out, not rpm. RPM can be high at very small throttle openings. (Less than 5%). without load. Nev

Sorry, absolutely right, I meant to add 'and run it under load', inadequate coffee intake for rational; thought...

 

 

[facthunter]

I know that you know. Nev

 

 

[Jaba-who]

I have to agree with all the sentiments expressed here. I would not like you to try that.

 

This is of a cobrahead modification we tried a few years back when we did a whole bunch of trials with baffles.

 

We tried this form of baffle on two aircraft. A corby starlet and a Jabiru 430.

 

The starlet had a 2200 with a limited room low radius curve in the scat tubing between the hot air mixer box and the carby. This type of mod fixed his problems completely.

 

On my 430 it was a little worse - not much but a little.

 

We put it down to my mates workmanship was much better. He had fine plates for the baffle, smoother edges wheras mine was a bit rough and agricultural. I made a guess his had more space and less baffle as a percentage of the whole cross section.

 

Maybe something like this, being more solid and unable to come apart, melt or get sucked into the carby or induction chamber might help the rest of us sleep better at night.

 

 

[facthunter]

What are you trying to achieve is the question. Surely it's mixing and maintaining a streamlined flow works against that. You are also increasing the wetted area. Something likely to cause a problem when you open the throttle quickly. The updraft set up on a Continental 0-200 seems like a fairly good arrangement. Nev

 

 

[jetjr]

I believe there is a slow swirl to air coming into carb

 

Its severity changes with airspeed/rpm/load

 

You often see two cylinders with quite high and the other quite low egt

 

One is getting fuel charge planned for the other

 

This spread difference can even move aroundand swap cylinders at different rpm.

 

Unless this egt spread is evened out you cannot jet the carb correctly and are constantly having too cool and too hot egt at th same time.

 

Yes a centre updraft would be better with mor even runners but hard to make this work on Jabiru, cowls, sump, cooler and exhaust all there.

 

 

[Oscar]

Becoming obsessed with egt spread, may not be the most productive exercise. Because of the differences in the exhaust runners, it is very difficult to ensure that you even have your egt probes measuring the same point on all four ( let alone six!) runners in relation to the standing wave pattern of the exhaust gas at any given combination of revs and load. Then, also, you have the problem of assuming that your egt probes are all registering with a similar accuracy - and who has test gear capable of checking at around 700C?

 

I've set up a calibration test rig for cht senders:

 

 

It uses a 100mm diameter x 300 mm or so long brass rod, heated by a stove element, into which we placed two laboratory-grade mercury thermometers. The cht sensors ( CAMit-supplied and from a highly-reputable supplier, not eBay specials!) were attached by tapped holes into the rod and the temps. at each sensor were also measured with a multi-meter temp. probe. The temps. at the test rig and the recording instrument were carried out with the complete reporting circuit to the instrument being that attached to the engine. It took nearly four hours to record both the rising and falling temps at 10C increments. The results showed that the cht sensors were reporting at the instrument, within 2%. And, we found that there were peaks and troughs in the accuracy throughout the range: you cannot just take one or two measurement points and assume that the reported temps. are linear with temp. changes.

 

A reliable, measurable 200C is easy to achieve in a calibration set-up; 700C is not. It's a very long bow guess to assume that any differential of your test of your egt probes in boiling water can be extrapolated to the same error at 700C.

 

A 2% error factor, at the critical cht's, is nearly 4C. Add in a bit of very minor deviation from 'best practice' flying, and you'd easily get 10C +/- of optimum for climb-out. And: that measurement is across four or six cylinders with identical temp. take-off points..

 

Now, let's think about realistic errors for egt's. ASSUMING all of your egt's are positioned absolutely symmetrically with regard to the standing wave patterns for the exhaust runners, a 2% reporting error at 700C is 35C. I've added egt's to my own Jab engine, and I would not dare to claim that they are absolutely symmetrical in position - and we tried damn hard to achieve that. I would not suggest that what gets reported is within 5% of actual.

 

When I built racing engines, we used 'plug-chop' to check the individual mixture to each cylinder, to tune our Weber DCOE carbies delivering one throat per cylinder through a straight (and very, very carefully matched) intake tubes to the cylinder head. 'Plug-chop' means a full-power run down the main straight and cutting off the ignition at full power and high load, so one could see the deposits on the plugs and identify if any cylinders were not operating at maximum combustion efficiency. We used colour-charts to determine the efficacy of combustion temperature to the optimum.

 

'Plug-chop' is NOT a good idea for air-cooled engines: it would introduce severe thermal shock. However, I suggest that inspection of the individual heads and piston crowns will give a fairly good indication as to whether your Jab. engine is running - in real-life conditions - reasonably, or badly. No matter what your egt instrument reporting is telling you, if all your cylinder heads and piston crowns have a uniform 'warm-caramel' colour with minimal deposits, then you are chasing your tail to try to get your reported egt's to all match. Your time would be better spent on improving the head cooling efficiency.

 

 

[jetjr]

Good info ........but

 

Ive used both dynon and CAE probes, see similar egt results, swapping a few probes around shows results remain with cylinder.

 

We are seeing 15-20% spread, so whilst 5% is significant, the spread is real and significant

 

To back this further cylinders running lean or rich show spark plug evidence as expected

 

 

[jetjr]

In respect to crown inspection, i now have borescope which can look at crow and valves

 

Got good images but aside from deposits not sure whats good or bad

 

 

[facthunter]

You have no idea what temperature the piston crown is running at It's just as important as the head temp. Nev

 

 

[Oscar]

Nev - In case you got the impression from my previous post that I think it doesn't matter, I would never argue that one should NOT have egt monitoring, and on all pots. Especially with our dodgy fuel quality if using Mogas, I believe it's an essential precaution to have any 'out-of-normal' situation reported as quickly as possible so one can take appropriate action.

 

My (probably badly-explained) contention is that with non-factory installations and instruments, it is going to be very difficult to be sure that the actual numbers you may be seeing on the instrument, are telling you the 'real' temps. that you are getting. With a fully-tested engine in proper testing conditions, where the variables can be virtually eliminated and the engine is torn down and inspected throughout the test process, it is entirely possible to determine that 'X' egt temp being reported by a specific egt set-up, is 'safe', or 'best' etc. - and for replications of that set-up, using TSOd componentry, you can have pretty good assurance that if you fly to the numbers you see, you are doing the right thing by your engine.

 

If you are doing your own aftermarket installation (as I have done for my engine), I don't think it is probable that you are going to get the level of uniformity of installation that guarantees the same reading from each sensor, just for starters. Those old Gunson 'colourtune' plugs (which are still available, I just saw) were very popular 'way back when'... and in fact, I know that they were used, with a fibre-optic line back to the cockpit, in testing the R.R.-built Continentals for use in the Victa Airtourer (and found problems which irked R.R., btw).

 

Just popping on a set of egt's and seeing some numbers come up that one doesn't like, doesn't necessarily mean that there is something radically wrong with your engine. Personally, if I had inspected the engine and seen decent evidence that on average, from real-life flying, the combustion appears to have been decent for all pots, I'd be more inclined to look at the numbers you are getting on the instrument as reporting a 'normal' situation and file that away as a reference point. Then, what you have is an 'alerting' set-up for sudden out-of-limit operation.

 

 

[Bruce Tuncks]

Thanks guys, I will be very careful to make sure that nothing can come loose and get sucked into the carby and that there will be enough power to take off properly. I will abort the take-off if it doesn't happen as fast as usual.

 

Oscar, I am guilty of just cobbling a system together, and yes I am a cheapskate, but I did try to get things equal between the cylinders. The probes are supposed to go " between 10 and 12 cm" from the flange, so mine are all at 11 cm. But the pipes have different bends so yes there can be a difference.

 

And I certainly take your point about calibration... ambient reads correctly and so did 100 C with boiling water, but this is a long way from the 800C that I am seeing on no.4. When I swapped the coolest probe with the hottest, the temperatures stayed with the cylinders.

 

Now I don't want to spend unnecessary time on this, but nearly 150 degrees (800 to 650) difference is surely too much. If that no4 can be brought down to 750 then I'll call it a day. Remember that the no4 has had exhaust valve-seat problems, and the exhaust valve seat would be the place where excessive combustion temperatures would show up, even with good cooling.

 

That's an impressive test-rig for CHT's. Is there nothing similar possible for EGT's? A pool of molten lead? Bugger that would be dangerous in itself, not to mention the fumes.

 

Well today is the day for some test flying, a bit of good luck would be nice.

 

 

[facthunter]

You are playing with a problem (fuel distribution) that has concerned many and is rarely fixed. In a critical engine it becomes more critical and necessary to have it right, for reliability.

 

On a slightly different but related matter

 

A backfire can do much more than dislodge an air cleaner. It could force a large amount of combustible gasses into another cylinder and presto there is broken studs or worse. Backfiring in gas engines causes shortened con rods and cylinder block breakages for the same reason.

 

IF you inject near the port there is not a manifold full of combustible air/fuel to cause the problem. It's all air up to that point, so isn't combustible. This is probably the best way to make this motor safe®. Nev

 

 

[Oscar]

Bruce: first up: absolutely all the best of British, mate!. Any glider pilot has a pretty good idea and experience of how to get it back on a small patch of ground safely, but I reckon I'm not the only one who'll be very happy to see your 'post-test' report! - to know that you're all good..and so is the wee beastie.

 

I absolutely DO understand why you'd be concerned at that degree of egt spread, and two things pointing to one cause on the one cylinder, is pretty convincing, I also agree. You've eliminated the possibility of serious probe error; as for position error, when we were putting ours on, we noted that using the recommended distance from the exhaust flange put the probes at different positions relative to the bend in the exhaust stack so we sort of tried to 'average' that out a bit, but whether we managed anything at all (for better or worse!) by that, I have no idea. We didn't have my engine on the test stand for long enough to run egt reporting tests (for various reasons), and since the exhaust stacks are is in the direct prop blast, it wouldn't be likely to be very accurate anyway.

 

If you want to get a fair idea of whether you're getting accurate (ish, even) reporting from your probes, maybe a lump of metal ( of large enough to be a useful thermal mass, but not so huge that you use up the energy resources of S.A. heating it!), with the probes making good contact, and an I.R. temp. instrument and a butane torch might get you somewhere. into the 500C range, anyway.

 

 

[Bruce Tuncks]

maybe a bit of progress, but not over yet... The plane had good power and didn't miss a beat but that no4 EGT is still too high.

 

Firstly, the 10mm tubes in the induction SCAT hose did nothing much. No noticeable loss of power but no improvement in the no4 EGT, so these were chucked out... bugger, they cost $2.

 

There is more work possible in this area of inflow to the carby. The problem could well be at the bottom end where something like the curved fitting shown by jaba-who looks like a good idea.

 

Secondly, the diverter-finger into the distribution box looks like it achieved about a 20 degree drop in no4 EGT, down from a max of 800 to a max of 780.

 

This finger was cut back a lot because of how Bex said it could break off, so what was tried was a tapered deflector 14mm at the base and 9mm at the top and curved out of the rear of no4 pipe into the distribution box. It was bent forward to intercept about 10mm into the 35mm distribution box.

 

This dropped the no4 EGT by about 20 degrees, not as much as I hoped from such a drastic measure . And I had the worry about it breaking off.

 

Then I disconnected the inflow SCAT hose from the external source to the filter box and let it suck from inside the cowl. This made no difference.

 

Finally, in some desperation, the main jet was drilled out from 2.11mm to 2.18mm. That is one size up in my box of number drills.

 

This appears to have dropped the ECG by 4 degrees.

 

The last take-off showed a max EGT of 276 degrees. which is better than the 800 C and about halfway to the goal of 250 degrees.

 

I don't know what to try next.. Based on what seems to have helped, a bigger diverter finger and an even bigger main jet might work.

 

 

[Oscar]

Well, that's $2 you won't see again - and it's good news that that's your major complaint!.

 

It might be worth a call to Ian Bent at CAMit; I know he has had some thoughts about the plenum design some time ago, but whether he's had any time to do anything with those thoughts, I don't know. Every engine made by CAmit, whtehre Jabiru or CAE, gets run for at least 20 minutes on his dyno, so if there's anybody who has seen a common pattern of egt spread, I reckon it'd be the lads up there. A dyno run doesn't exactly replicate the installed configuration, but if the problem is downstream of the carby rather than from the air being delivered to the carby, I'd think it would show up.

 

 

[Bruce Tuncks]

Here's the flow straighteners I have tried, and neither reduced the EGT on no4. It looks like I need to make a proper one like jaba-who did.

 

 

 

[facthunter]

I don't think straightening the flow is the issue. It probably needs maximum turbulence before the first split. Nev

 

 

[Oscar]

Nev: straightening the flow is AN issue, but it's not the only one. The early Jab. airboxes are bloody crude - mine, which was developed as a test mule for the original 2200, is so bad, it's beyond appalling; the Jab. installation manual has some good advice for cleaning it up reasonably, but the convolutions from the airbox to the carby as a result of the tight installation, remain as a problem.

 

However, the position of the butterfly also seriously affects the airflow downstream of the carby, into a plenum that is extremely tight. That is minimised at WFO; Bruce's experience indicates to me that at lower openings (cruise power) he's getting quite serious pressure differentials within the length of the plenum, and the swirl is introducing a lateral component to those that starves the #4 pot. I suspect that the combination of swirl and butterfly position, is the reason that tilting the bing on a Jab. can make a significant difference to egt spread.

 

 

[facthunter]

Oscar, Radials get over this by having a diffuser which is easily further developed into a supercharger achieving two things at once. Straightening the flow will reduce mixing as it can remain stratified. Corners make the heavier part of the charge go to the outside as well. Having great long manifolds full of combustible mixture is dangerous as a backfire can destroy the motor by severe overcharging one cylinder. If someone runs a motor on test with Autogas this will prove or disprove my theory. This is surely quite easy to arrange. Modern fuels don't vaporise easily and are really more suitable to injection. Nev

 

 

[Jaba-who]

Here's the flow straighteners I have tried, and neither reduced the EGT on no4. It looks like I need to make a proper one like jaba-who did.

If it didn't help with the straws - is there anything to be gained by going to the effort of making a more formal one?

 

If it had worked with straws then I'd be thinking -" yep go ahead and make a safer one." But not sure it's going to help.

 

I guess it almost certainly won't hurt.

 

 

[Bruce Tuncks]

I like your comment jaba-who.

 

Today I straightened that finger which pokes from no4 pipe into the inlet . I did have it bent over so the tip was axial with the distribution box. Today I straightened it so the tip was about 45 degrees to the axis, and it scooped in about 5mm more on account of extending further.

 

Well the maximum EGT is now 754 C.

 

The max readings were: No1=677, 2=640,3=734 and 4=754 at take-off max throttle. Max diff=124 degrees.

 

Cruise readings were:

 

1=645; 2=644; 3=672; 4=701 at 2800rpm and 90 knots. Max diff= 57 degrees

 

So I'm about ready to call it a day, but I'm still worried about that finger coming off and spoiling a flight.

 

 

[Bruce Tuncks]

Physical reality sure is hard stuff to deal with.

 

Looking at the temperature of 754 C for no4 and the 640 for no2 which is just behind no4 as far as the fuel mix flows, I decided to have another go.

 

Firstly, the main jet was increased to 2.32mm bore ( n0232).

 

Next, the finger-deflector was replaced by a slant-cut cylinder as a flow interceptor.

 

If you imagine the pipe to no4 cylinder protruding into the distribution box by 18mm and then the pipe front cut away on the slant you get the idea. No more pussy-footing around thought I.

 

The idea was to intercept the droplets of fuel which were "clearly" bypassing the no4 exit and going on to no2. So this mod should surely richen no4 at the expense of no2. Plus there would be no finger-like protrusion to worry about if it snapped off.

 

So after a few hours of work, here is what I measured:

 

Max readings: No4=710, no3=752; no2=633; no1=698

 

So the interceptor richened no4, but at the expense of no3 not no2! But no3 is opposite no4 , and no2 is behind no4. Swirl is the only explanation I guess; but its hard to visualize what is happening .

 

On cruise no4=690; no3=660; no2=620; no1=687. This is even enough, 70 C difference between hottest and coolest.

 

You would be excused for thinking there might be an unacceptable power loss and I was looking for this but didn't notice anything.

 

Time for a rest. I reckon the engine is better now than at the start so its ok to just use it like this.... or maybe try a bit of carby tilt next.

 

 

[facthunter]

I'm glad you've given the straws away. Some people have been known to clutch at them which doesn't appear to help. Nev

 

 

[Oscar]

Sounds like a fairly good result!. You MAY find that there are some quirky spots in the rev/load matrix because of that rather large intrusion to the plenum, but with full monitoring, you'll get the information you need to be aware of those. I presume you have looked at the Jab. 2200 installation manual re the advice they give for ensuring the airbox is operating as best as can be tweaked? Some of the early ones were very, very rough..