Arrow Engine Failure

[NoWingsAttached]

I remember a dirt bike racer telling me the 2-cycle motors ran just as well backwards as forwards, and that it was not uncommon for a 2-c to pre-fire and start backwards.

My very first personal encounter with ANY 2-cy engine other than a leaf blower was an Arrow GT 1000cc, 110-120 hp, ported-piston, Italian engine, out of production. It started backwards anytime the prop was free-wheeling on the sprag and you hit the start button. I never gave it much thought, as I thought that was "normal". I flew it 10 hours without a hiccup.

After sitting unused for about 18 months, I flew it for 3 hours and noticed some periods where it lost power under load, then the engine locked up during a touch-and-go @ about 20 ft up. Some of the early issues were fuel delivery. But those solved, she locked up from something different. I thought a rod bearing disintegrated due to some failure of lubrication and corrosion. But after a thorough autopsy she showed no signs of rust at all on the needle bearings for the connecting rods on the crank. I have always been at a loss to explain it, and it has bugged me. I think I've found the answer.

The Arrow engine I bought to replace it with never started backwards, and never had an issue with starting at all. I decided to check everything over thoroughly and not take anything for granted as I did before. After going through many things I realized I haven't had a working timing light in decades. Hmmm...which means I NEVER checked the timing on the blown Arrow i bought originally, which only had 13 hours on it since I bought it, with 50 hours on it total, before it locked up.

I got a new timing light, checked the motor and the new Arrow is dead nuts on.

I looked at the static location of the pickups on the first Arrow. The timing is 28 frikking degrees advanced, almost TWICE what it is supposed to be. Damn, am I an idiot for not spending $75 bucks on a new timing light 2 years ago and checking it or what? Looks to be a case of a crap job on engine tuning and set-up on the poor little Arrow, not a failure of the bearings nor neglect. They were merely collateral damage, not the culprit.

Also turns out that the newer, lower HP, 1000cc Arrow I now have uses a shorter gear, 1:2.77, not 1:2.58 as on the beefier GT motor. So I put the 68" Warp that came with that setup on her, and it brought the max revs down to below redline. The old prop and pitch settings allowed the motor to over-rev.

You know, the prop that came with the first GT Arrow is only a 100hp Warp Drive, but the one that came with Ira's engine is a high-horsepower Warp Drive, meaning.....

I can hook it up to the Yamaha!

And...I have been thinking...that sprag clutch drive is beefy enough to take the hp of the Yamaha. Gearing goes to 1:3.01 on this outfit. The Yamaha turns 8700 at the shaft. Hmmm... I need to get my calculator to see how this works...

Engine redline is in the 10k range, at the output (reduction) shaft it is 8700, and at 1:3.01 that gets us 2890. 2850 prop rpm is then 8579 shaft rpm. Hmmm. This is suddenly much easier - and lighter! - than the drive shaft and CVT clutch design I was moving towards. I simply swap gears on an existing redrive/prop combo I ALREADY have, and presto. I mean, if a sprag clutch solution to damping eng-prop pulsation on this 4-cylinder, 4 cycle, 150 hp, 125 LB, Middle-Heavy weight bad boy is good enough for Neil, why not good enough for me?

He was pretty darn happy with the sprag solution when I talked to him at Bensen Day's. You know, I am probably going to put the Yamaha on Ira's low-rider frame. Yep. No, I am not crazy. There are OTHER plans for this machine in it's final revision that you will never guess at. Something I think is 100% unique. Safe, and LOADS of fun. For what I am about to do, a low rider is not only acceptable, it is the ONLY way to go.

Can you guess my intentions?

[CLS447]

Are you saying that both of your Arrows are now junk ?

[NoWingsAttached]

Oh, goodness no. The second one is fine. I am flying it a half hour out, half hour back, landing, checking things over and then flying again for another hour...my wife is starting to feel the gyro widow pains, so i need to give her some extra TLC tonight...

[GyroCFI]

I put hundreds of hours on the Arrow 1000GT first generation back in the early 90's. I did notice whenever it backfired when I was starting it, the sprague clutch would flip one of the dogbones which would lock the sprague. I got really good at replacing that sprague clutch and actually kept several on hand. you could still fly with it like that but I never did except one time when I could feel the change after I was on an extended glide and kicked the throttle too hard, so I knew it locked. I made it back without incident and really the only thing that was out of the normal was when I shut down the engine, the prop came to an abrupt stop with a shudder.

I had more problems with head gaskets blowing and having to helicoil the head studs in the cylinders on the exhaust side of the engine.

[NoWingsAttached]

Originally Posted by GyroCFI

I put hundreds of hours on the Arrow 1000GT first generation back in the early 90's. I did notice whenever it backfired when I was starting it, the sprague clutch would flip one of the dogbones which would lock the sprague. I got really good at replacing that sprague clutch and actually kept several on hand. you could still fly with it like that but I never did except one time when I could feel the change after I was on an extended glide and kicked the throttle too hard, so I knew it locked. I made it back without incident and really the only thing that was out of the normal was when I shut down the engine, the prop came to an abrupt stop with a shudder.

I had more problems with head gaskets blowing and having to helicoil the head studs in the cylinders on the exhaust side of the engine.

I have kind of arbitrarily decided to torque the head bolts to 25 pounds. That is right in the middle of maximum for the bolt grade, and where leakage seems to start at the gasket. The idea is to check them every 5-10 hours to make sure they aren't loosening up from the heat cycling and the vibration, since they aren't torqued to max, and save pulling threads out of the aluminum.

Does this make a sound plan?

[GyroCFI]

Originally Posted by NoWingsAttached

I have kind of arbitrarily decided to torque the head bolts to 25 pounds. That is right in the middle of maximum for the bolt grade, and where leakage seems to start at the gasket. The idea is to check them every 5-10 hours to make sure they aren't loosening up from the heat cycling and the vibration, since they aren't torqued to max, and save pulling threads out of the aluminum.

Does this make a sound plan?

I did the same thing, I don't recall what torque I put on the bolts, but it was less than the recommended. The problem was that in my opinion the cylinders were made of a lesser quality metal than they should have been, almost like pot metal. Eventually every one of the lower three head studs would pull out of the cylinder. I would helicoil them back in but that repair was not permanent, i eventually had the helicoils pull out too. I replaced all the cylinders twice during the time I owned that engine and even put on a cooling shroud to help get more air to help cool them down per the factory recomendation. It did help to push out the failures longer, but they still eventually failed again. I sold it with a brand new set of cylinders and gaskets all around, I don't know how the owner faired with the engine after that.

[NoWingsAttached]

I made an error, I only use 15 ft lbs of torque, not 25.

I think I may have a rather simple solution to the problem of pulled head studs. The length of the stud inserted into the cyclinder is 16mm, the deepest you can get with the stock hole as drilled in the cylinder jug. The thread pitch is 8 per cm, so about 12 effective threads in the hole are engaged. For easy math let's say we torque the bolts to 12 FP, and thus the total force spread over the 12 threads is 1 fp per thread.

Guess what? The available material on the cylinder allows for a hole twice as deep, 32mm! By drilling it all the way out and doubling the depth of the hole and length of the stud, I can better than halve the load per thread! That would be equivalent to 7 FP on the existing studs!!

Last Saturday after a decent pre-flight check, I went up for a quick warm-up lap around the pattern and returned to ground check everything again. I usually don't do this, but I had planned on using the front seat tank for the first time in 4 years, so I wanted to be absolutely certain nothing happened that would be due to the new fuel load in the fore tank.

I'd filled it with gas, let it sit for a week, then drained that and put it in my Durango. Then I loaded the tank with fresh oil-gas for flight. During the second inspection I noticed some oil on my prop that wasn't there at first. I checked the heads and found a leaky gasket on the rear set of jugs. The head bolts on the cylinders closest to the prop were mostly all loosened up. The ones forward were fine.

I re-torqued every nut on the studs to 15 fp, but the one at the point of leakage felt mushy as I tightened it, so I quit turning it and left it around 13-14fp. This was a lower stud near the exhaust port, such as Don talks about here. I went flying again, and the gasket continued to leak a bit. Since this stud feels like it is mushing out, I should probably just go ahead and helicoil it - all the way to the maximum depth the available material allows, not just the same depth it is now. While I am at it, I should probably drill and tap all of the studs to maximum depth - before they need to be helicoiled.

I know I over revved the motor once or twice, briefly. I think that may have been what lead to this pulled stud. How did I do that? The notes I had on the redrive showed that it was 1:2.58. I was using that redrive and a prop pitched for a 1:2.58 ratio combination. When I noticed the tach going well past redline, I verified first the tach, then the gear ratio and found out it was in fact 1:2.77.


One other thing: When I did the second ground check and found the leaky gasket, I also found a drop of fuel on the bottom of the fuel filter that wasn't there on the first inspection as well. It is the type that screws together, with a removeable screen insert. As I tried to tighten it up it came apart in my hand. The threads had stripped out. I had to run up to the store and get a replacement. Imagine if THAT had failed in flight. Two potential engine outs wrapped up in one, single, 10-minute flight!!! You know, i usually do my pre-flight, then take off and fly until I am at the reserve level of fuel. I rarely fly 10 minutes, then return for a second ground check like this.

Either I am really lucky, or I am really unlucky.

I sometimes leave the radio on when I am working on the gyro to listen to traffic. At the same time all this was going on, a Cherokee radioed in that he couldn't tell if his gear was down or not.

Damn gremlins were everywhere last Saturday.

[GyroCFI]

I thought about drilling and pinning the stud bolt through the side of the cylinder where it is built out but never got around to doing it before I sold my machine. I think the metal just gets weak and turns soft on the hot side of the engine.

[NoWingsAttached]

You know...if there is enough meat to gouge out of the aluminum to accomodate a helicoil, why not cut below the boss to accomodate the use of a skinny nut to hold the stud, after drilling through the other end of the pocket as described above? Jeez, with that you could torque the heads to 25 fp and rev the motor to 8000rpms.

[CLS447]

Your wife ? You got married again ?

The one with the kitchen counters ?

[GyroCFI]

Originally Posted by NoWingsAttached

You know...if there is enough meat to gouge out of the aluminum to accomodate a helicoil, why not cut below the boss to accomodate the use of a skinny nut to hold the stud, after drilling through the other end of the pocket as described above? Jeez, with that you could torque the heads to 25 fp and rev the motor to 8000rpms.

Good Idea! I looked at all sorts of remedies for that problem. Sold the gyro before trying to implement any of them.

[NoWingsAttached]

Originally Posted by CLS447

Your wife ? You got married again ?

The one with the kitchen counters ?

Wow, you are way behind Chris! LOL. Nope, not the beauty who posed for those shots in the photos I posted Summer '09 (time flies!) here that everyone liked so well. This is a different lady I met on a business trip to Hong Kong in 2008. We got married 12/4/09, a while after I broke it off with the kitchen counter lady (she was talking about wanting a baby, and I felt I didn't want to start from scratch with a brand new baby at my age), and then she moved here 10/6/10.

[NoWingsAttached]

Originally Posted by GyroCFI

Good Idea! I looked at all sorts of remedies for that problem. Sold the gyro before trying to implement any of them.

Your idea about drilling and pinning might work, if I understand right you would use a little roll pin through the stud?

[CLS447]

Wow Greg, congrats ! Where does the years go ?

[GyroCFI]

Originally Posted by NoWingsAttached

Your idea about drilling and pinning might work, if I understand right you would use a little roll pin through the stud?

Yep, that's what I was thinking. Never had a chance to try it though.