Taxi Tests

This is taxi run #11 I think, who’s counting anyway?  I wanted to try and sort out the radio problem AND see how the cooling goes with the cowls on. I’ve just added the heat shields, the top cowl stiffening works a treat and I was expecting a good result and even start runway tests the same day. On to first flight in a week.

Not so…

Previously at low RPM, cowls off I could maintain temperatures on the ground. This is my big goal for the plane so I can sit in a queue at airports for as long as I need without shutting down the engine because it got too hot.

Yes I could then use a handheld and do a hot start as required but this is a pretty awful way to operate. Its a lot of extra pilot stress too and I’d be hot on the take-off roll when I finally got there. There may also be problems restarting the engine when it was my turn. I’d have to restart each time the queue moved too.

Cowls on and my cooling is an epic FAIL. Even at 860rpm the temperatures kept slowly going up. I got to 212F after the radiators although the temps before the radiators where not much better. I have three sensors for all this. I was also up to 13psi coolant pressure when I pulled the plug and taxied back to the hangar.

The cowls were REALLY hot to the touch when I got out. Oh and no joy on the radio problem either. Anyway that puts a big stopper on my imminent first flight. I could go ahead and fly, but then what? It’s going to be far easier to solve this problem at Moorabbin that relocating to Leongatha where resources are only what I bring in the car and things like wiring would not happen easily. Given it just doesn’t seen to cool down quickly after the engine is stopped that idea of shutting down and restarting in a queue is not really viable either.

What to do? How to fix this ground cooling problem?

Get data. I need to do a lot of testing, timing how long it takes to get hot. First cowls off, then bottom cowl on with top cowl off then bottom on top cowl off. I need the numbers. I am expecting to find it OK bottom on, that would tell me that I need a huge vent in the top cowl. What I expect might not be what the data tells me.

Get data. Fans do not work properly without shrouds. Testing last night with bits of paper revealed that air is getting pulled in from the HOT side into the fan at the sides. As a quick and dirty test I need to block off the areas around the fans at the back of the radiators and see if this improves my test numbers. If it does, I am looking to add a third fan and test this too.

If that works I have an idea for louvers at the back of the fans which are closed on the ground but open in flight. This is a common automotive practice that someone smarter than me pointed out. So if the blocking off thing works I have a plan to make it useable.

If all this fails, what comes next is ugly. The tests should tell me if I need to vent just the top cowl or both top and bottom. How big to make the vents? How to I find out without cutting up the existing ones and cutting too little or too much or in the wrong places??

Yes I’m going to have to make a test cowl or even several cowls. I’d use the existing ones as a positive plug (mould). No small task. It doesn’t have to be flight worthy, just strong enough to sit in place while I taxi around slowly. It will probably have to be three ply of carbon too so I can get it to hold shape as I cut bits out. What comes off could be re-joined as I get the shape and size worked out.

If big vents work I am going to have to have doors that open with linear actuators and they don’t like heat at all. Remotely locating one would be very challenging and if its via cable it may need to be a push pull. A real can of worms, maybe something like the gear doors that spring open and are then held closed. Essentially I’m looking at cowl flaps that some GA planes have as standard. Maybe I can rig up a manual cable and set it so the airflow will tend to keep it closed.

Secondary to this is to get more control over the water pumps and fans. At the moment it is automated so the fans and pump at full power kick in at 170F or whatever I set it at. I do have an aux pump on a manual switch. However at the moment it is only circulating coolant that has not been cooled well at the radiators so it does nothing much at all. That’s the blocking off thing I need to look at. The other thing is to may have the third fan also just an on/off so I can at least get the fan and pumps going early if I need to. It would be on me to get them off once I’m in flight. This might help stop the heat build-up that seems hard to get rid off.

It looks like I’m not flying this year after all. That leaves questions on my pilot currency for first flight as I want to let the old plane go by the end of the year. I have decisions to make in a week or two.

This is the 10th taxi test now. Still just around the parking area and not yet on the runways. The cowls are still off as I evaluate the cooling, radios and engine further.

I’m now using a ‘fuel map’ from another LIO-390 engine what has the same Superior cold air induction. I did try the map from the test cell for this engine but it had the Skydynamics sump at that time. First up, the fuel map that was with the plane when I received it was unusable and the cause of many expensive and serious problems. I do not believe I altered it by accident, I don’t even think that is possible given it was so far wrong.

The current map is usable but nowhere near right. The idea is to have an air/fuel ratio (AFR) at 14.4-15 for cruise and 13.2 under load. So I need close to those numbers where I read the AFR in the cockpit over the whole rpm range. You can download the file to a PC, adjust each parameter for each 100rpm and each map setting manually and then upload it again.

I can tweak in the cockpit  with the ‘Tune’ function but not so precisely. It takes the numbers and averages them. Another way is to see what I have and use the Fuel Trim knob (FTK) to tweak on the fly. Those changes aren’t permanent and as I change the rpm or the MAP changes in the climb or descent, I’d have to readjust the FTK again. A bit like a mixture control so no big deal.

For now what I have is useable but a little rich over the range. I’m still having problems starting the engine. This may be the rich/lean settings at very low RPM  or it may just be me learning what the engine likes in terms of using the accelerator pump which is a fast throttle movement.

The last run I got the engine to 2300RPM in the runup bay. The brakes held but I really stood on them. This is close to what I’ll be using for take-off so we are good here. However as expected the engine got really hot as did the coolant and oil. The coolant pressure was right up close to maximum before the overflow is engaged as well. I backed right off of course, warnings were on the panel and I got a nice female voice in my ear telling me that certain pressures and temperatures were in the red. Not in those words however.

Despite returning to idle levels the oil stayed pretty hot, the coolant levels and pressure came down slowly. What I have found is that if I stay at around 1000RPM or less I can sit there without the temperatures rising. The two fans are doing the job given we have no air through the radiators from forward motion. I looked at the data as there is room for another fan which I thought might be needed with the cowls on. However with the fans and two pumps running, current draw gets to 54 amps and that is almost the limit for the 60amp alternator. The 5.2″ Spal fans draw a maximum of 5 amps so I’d be getting very near the limit if not over it at low RPM. We might be able to tweak the regulators up a fraction however. Without the fans, one pump hardly running and flight instruments and lights I’m at 34 amps.

I have some heat shielding for the cowls on order. I need to adjust a fastener and a few other configuration jobs. I’m still trying to sort or the squelch levels on both radios. They break through with the engine on, its perfect engine off so I need to tweak that while its running. To do this I need ‘config mode’ and to turn off the engine instruments. The GTN750 radio can be done OK instruments on. This is my next task in the runup bay. I will put the cowls on for that but I’ll keep things at 1000rpm so nothing gets too hot.

Once the radios are sorted out AND the heat shielding on the cowls is installed I can contact the tower and see when I can use the active runways for slow and then fast taxi.

So far, four taxi tests of 15-20 minutes each have been completed. These were just slowly around the parking area and runup bay, getting a feel for the plane. I’ve also calibrated the magnetometer so the ADAHRS has stopped showing error messages.

I have been at the front of the hangar for a change so able to get the plane out. The weather doesn’t matter for these tests.

The last test was with the cowls on. When the engine starts I get a general coolant pump warning, low coolant pressure warning, low pressure from the oil and so on. This tends to inoculate a person to these messages and given I’m not above 1300rpm I pretty much ignored them.  The pump warning did go off and come on again so I turned on the backup water pump just to be sure.

When I stopped the engine back at the hangar there was a LOT of coolant on the ground. It turns out I’d blown a Welsh plug! This was the one we replaced that was leaking from when the engine was delivered. Clearly the installation technique needs improvement.

Of course I was not going to get of lightly. This failure in the air would mean I’d have to shut down the engine or wait for it to seize. On the ground it meant the engine just got hot… but the coolant temps were constant as there was nothing to show the sensors with no water present. I didn’t know. Given the water pressure is zero at start-up with a warning, then it slowly climbs towards 15psi. I missed any message that it must have dropped from that.

So the engine got hot and the cowls hot enough to blister the paint both top and bottom. <sigh> At some point these cowls will need a complete repaint. You can see the marks above like a case of the hives. Review of the engine logs showed nothing catastrophic so I really got of lightly if you consider the big picture.

These glorified bottle tops are cheap but losing an engine isn’t.

I googled how to put them in, I watched five videos and each one had its own method. After a bit I understood what to do rather than follow the instructions like a robot. Every single person also had a favourite sealant too. The more I looked, the more options were presented.

I went for the anaerobic sealant I have used on the water system sensors. This has worked well where others have leaked, so more of the same seems a good thing to try.   The important part seems to be getting the plug in via the edges and not touching the middle. The wedging holds the plug in under pressure, not the sealant.

With that sorted out, the next day I drained the coolant which was only about 2 litres despite the sight tube level indicator showing full!  I filled again with the 50:50 fresh green stuff by pulling a vacuum at the cap this gives me a much better chance of having no air pockets in the liquid.

What to do about these CAS messages that are false on start-up? It turns out that with recent software changes to the G3X you can make messages ‘conditional’. We have set things up so many of these messages happen after the engine has been started and running for a bit. This should make them appear when they are REAL warnings, not false ones so I’ll take more notice.

We also reconfigured the fuel level system to now display on the G3X rather than the AG6 annunciator button that I had spent a huge amount of time and money on. When taxi testing it seems stuck on low sump despite there being 60 litres of fuel in the plane. It’s just been a PIA to program and hook up.  With software updates to the Garmin G3X it can now  accept different voltage levels for fuel annunciations.

Here’s the AG6 button which is currently disconnected and dark. I’m thinking maybe it could be a Master warning for all the major CAS annunciations but I’m reluctant to go to my long suffering avionics guy right at the moment!

Next up is getting the exhaust pipes tweaked again so the sit in the middle of the cowl openings. This is a cut and shut procedure. Then its back to the engine fuel tuning issues which are not yet resolved. I also have to stiffen the top cowl. I’ll do a separate blog entry on all that. Its a bit of work.

My Long-EZ build moved under its own power for the first time, November 1st 2024. It has been 12 years and 11 months since this marathon began and over 16,000 hours of logged build time.

We had finally found the cause of the latest engine problem which was the fuel mapping and with a little difficulty the engine started.

Here’s a link on my YouTube channel with slightly better resolution.

Since then I’ve done 4 runs in total of around 18 minutes each. The steering was quite stiff which was a concern but that only needed some lubrication on the nose gear and now it really turns well and feels much better in the taxi. It is extremely sensitive so I need to be on my game on the runway.

I also found that with about 2mm off the nose gear cover and one AN size shorter bolt down there with a washer removed I can now rotate the nose wheel 360 degrees on the ground. This makes pushing it backwards where the nose must rotate 180 degrees much easier.

The cowls are not on yet so that will effect the temperatures in a negative way. We are starting on a cut and shut for the exhaust pipes today. The ends are too near the top cowl for my liking and I have found a gun welder to do the job.

The temperatures hasn’t reach maximum at low RPM after 20 minutes so that’s going in the right direction. However the Delta T is only 6 degrees. I need this higher. That’s the difference between the coolant temp from the engine and then measured after the radiators before it returns to the engine.

The only control I have with the Coolant temperatures in the cockpit is a switch to turn on the second water pump. This is why I will try setting the fans a bit lower for when they cut in. Mind you, we have no airflow over the radiators really at taxi speeds so its just when the fans cut in and I don’t know how effective they are yet.

There is an issue with the rudders not returning to fully closed. This is easily fixed by pulling the pedals forwards in the cockpit with my feet. The problem is that the angle from the pedals to the cable run is too severe to allow the rudder springs to pull back into line. It is likely that the airflow in flight will fix this but I will have a look at dealing with the angle involved. Its a bit of work and would need a new bracket on the pedals a bit higher and perfectly set up so I don’t have to alter the cable length. I wish I’d picked this up earlier, I thought it was paint too thick on the rudders preventing full closure. At least I know exactly what’s going on now.

There is also an issue with the fuel level electronics giving a false indication, so that’s another one. I need to adjust the squelch levels in software too. Its always open which is really annoying! That needs a borrowed laptop unfortunately. My old one is pretty much toast.