Fearless Garage

The fourth rally in the 2011 MCO Open Road series was held Wednesday evening and started at the Tim Horton’s in Stittsville. Gary and I met at the Cheshire Cat, where the rally would end, so he could leave his van there. When we got to Timmie’s a couple of kilometres away, there was a small group of rallyists and the usual assortment of volunteers to man the checkpoints. We were the third team to register, so we were designated Car No. 3 and we would depart at 8:03 PM.

The first section was simple distance to turns up to the odometer check, followed by a single average speed for another 10 km or so to the end of section. We had no trouble with this part and had some time in hand before beginning the second section, allowing Gary the luxury of working ahead on calculations while parked by the side of the road.

Section 2 was a series of distance to turn instructions which were completely out of sequence. So Gary had to be careful not to miss one, thereby causing us to stray off course. This section also covered the most challenging road of the rally – the Old Perth Road. Glen had told us at the drivers’ meeting that when he greened the rally the day before, the condition of the roads had been awful, with a mixture of mud and slush that was very tricky at speed. However, when we drove it the ground was essentially dry and somewhat frozen, so the traction was quite good. The speed on this road was 65 km/h and it was challenging to maintain. But with excellent visibility from having corrected the aim of the driving lights, it was reasonably comfortable negotiating the numerous hills, crests and hidden turns. The potholes and dips were another matter though, as some sections were extremely rough, including one large dip that caused both of us to graze the ceiling.

The next section included another dirt road and a few paved transit sections and the instructions were simple tulip diagrams with numerous speed changes. Now it had become a leisurely drive in the country on a clear night and was quite enjoyable. The fourth section had a table of mixed up directions which had to be sorted and put in order, but the roads were pretty good and the transits were long, so we had no trouble staying on route or on schedule. The last section had a new twist in its instructions – the distances started at 14.7 km and ended at zero! So I reset the odometer to 14.7 km at the beginning of section and adjusted it to run backwards, so it counted down to zero as we approached the finish. This disabled the average speed function on the computer, so I just drove at the speed limit and Gary calculated our arrival times at various points along the way.

Our scores at the checkpoints were all pretty good, except for the first one, where we were about 45 seconds late. We had assumed that the average speed at the beginning of Section 2 was the same as the finishing speed for Section 1, but this turned out to be a mistake. In fact, I think it was meant to be the average speed for the whole rally, which was given on the last page of the instructions! At the other controls we had scores of 0.2 (twice), 0.1 and zero. At the finish we got caught by the above-mentioned overall average speed issue, so we checked in two minutes earlier than we should.

We finished with a score of 3.2 over the entire 112 kms, which could have been 0.5 if we’d figured out the overall average speed trick correctly. But it was still good enough for first place in Expert class, so we’re not complaining. There were only seven teams competing, but at least there was one completely new team in a new Subaru. They made it back safely and had a great time.

Video is available at:  http://www.youtube.com/watch?v=K7g5EynpfnQ

After my Porsche engine blew up at Road America in early September 2010, the mechanic who had built it – I’ll call him Chuck, not his real name – tore it down and we discovered total spark plug failure in cylinders 1 and 3. The ceramic insulator in #3 had broken and slid down the anode, allowing the spark to arc over to the cylinder wall. The result was a hole of about ¼” diameter in that cylinder.

After talking to numerous expert engine builders all over North America, I conclude that the reason for the failure was based on all of the following factors:

1. The timing was advanced too far – at 35 degrees BTDC – for prolonged running on 91 octane fuel. It should have been about 29-30 degrees.
2. The air/fuel mixture was set too lean for an air-cooled engine to run almost exclusively on the track, 13.2 at 6000 rpm. It should have been about 12.5 to 12.9.
3. The fuel I was using should have had a higher octane than 91 – at least 96 if not higher, especially for the timing we were using.
4. I should have been using colder spark plugs, to disperse more of the heat.

After collecting all of these inputs, Chuck was not pleased. He didn’t feel that many, if any, of them were credible. And he hated having his work and his opinions questioned. We had quite a debate about which spark plugs to use, as a result. He was relying on advice from PMO regarding the timing and plugs and advice from Motec on the air/fuel mixture. I don’t think he’d ever built an engine for almost exclusive track use, other than his own. And he has a reputation for breaking his cars before their engines can fail!

Meanwhile, I had reached a further conclusion as to why the left side had failed, while the right side still seemed healthy. I had noticed before going to Road America that the hose connecting the left side heat exchanger to the heater control box was disconnected. The hose was in bad shape and I wouldn’t need cabin heat at Road America, so I didn’t bother reconnecting it. But the result of this was that much-needed cooling air for the left side cylinders had an easy escape route out of the heat exchanger. This would make enough of a difference to explain why one side failed before the other. One of the expert engine builders agreed with me as soon as I told him this. Chuck was not particularly impressed by this when I told him.

I ordered new JE pistons and EBS cylinders, as well as rings and gaskets. Chuck reassembled the engine with these components and restored the original jets in the PMO carburetors. I told him I wanted the timing to be set at 30 BTDC and he did that. When I drove out there with my trailer at the end of October to pick up the car, the first thing he said was “It won’t run.” Initially I thought he was giving me a status report, but he was actually expressing an opinion based on his frustration.

I had asked him to put in the coldest plugs (NGK BP9ES) and it was very difficult to start, with a lot of sputtering and backfiring at idle. He set the timing to 30 BTDC and then said “It’s yours. There’s no bill. You and I are done. I don’t build stuff based on what you read on the Internet.” And stalked off. That’s the last time I’ve seen him and it’s no great loss.

Here’s what’s happened and what I’ve done since getting the car home.

1. I verified that the correct main jets and air correctors had been installed in the carburetors.
2. I installed the hose connecting the fan shroud to the heat exchanger, which Chuck had not done. This is rather ironic, since the lack of cooling air had been the cause of part of the failure.
3. I installed street brake pads to replace the track pads which had been on the car for Road America.
4. I checked the tightness of all oil line connections and the oil strainer plate.
5. I installed hotter spark plugs (NGK BP7ES), to reduce the amount of backfiring.
6. On a short drive, the air/fuel ratio (AFR), was about 11.5 at 4500 rpm at wide open throttle (WOT).
7. I connected both heat exchangers to the heater control valves under the rear seat.
8. After a couple of short break-in drives, I installed even hotter plugs as a temporary measure to reduce the backfiring and allow the engine to run smoother.
9. On one break-in drive the engine seemed to be running well, although at idle it sounded like only five cylinders. At one point there was an enormous backfire, which gave me great concern.  After getting home I determined that the number 4 spark plug wire had not been properly connected at the distributor end, so it wasn’t firing. That would explain the backfire, as well as the wet condition of the plug.
10. A neighbour told me the car was smoking all the way down the street. So I checked the compression and did a leakdown test. All cylinders were between 105-125 psi, which is low but perhaps to be expected, since the rings wouldn’t have seated yet. Leakdown was 13-18%.
11. I cleaned all of the carburetor jets – both main and idle – as well as the air correctors. There was some debris in at least one.
12. I adjusted the intake valves, finding that all were a little loose relative to the spec of 0.004” cold.
13. I turned all of the idle mixture screws out ¼ turn to make it a little richer. This smoothed the idle and reduced the backfiring, as suggested by the helpful neighbour.
14. I cleaned and re-gapped all of the plugs, as well as cleaning the distributor contacts and verifying good connections for the plug wires.
15. I took a couple of long break-in drives, to reach the first 1000 km. The engine was running strongly.
16. I changed to regular oil, from the break-in oil, adjusted the exhaust valves and tightened the rockers.
17. I took one more long break-in drive and the engine was running well.
18. I installed two cylinder head temperature (CHT), gauges, using thermocouples on plugs #3 and 6,  reasoning that they would receive the least benefit from the fan’s cooling air.
19. I removed the oil reservoir, to adjust its float in order to get correct readings on the gauge. While there, I made a new support for the right rear fender, to replace a section that had rusted out.
20. I adjusted the timing to 5 BTDC and then 10 degrees, to smooth out the idle as a temporary measure, but it was still rough. Then I discovered that at some point I had removed #2 plug and not reinstalled it! Of course the engine ran much better with it connected.
21. I had been curious about the location of the number 1 plug wire when I was working on the distributor, so I checked some old photographs. Sure enough, it used to be positioned around 4 o’clock on the top of the distributor, but now it was more like 3 o’clock. And the adjustment slot was at the extreme end of its play, preventing me from retarding the timing easily. I removed the distributor and found that the rotor was pointing about 30 degrees before the notch on the distributor’s base. I rotated the distributor clockwise by one tooth (30 degrees), and restored it to its original orientation. Now the rotor lined up well with the notch, but the adjustment slot was at the opposite end of its travel. At least now I could easily retard the timing, but it would be hard to advance it too far – which was fine. I set the timing to about 5 BTDC, which gave 30 degrees at 3000 rpm and above.
22. At this point I wanted to try the colder plugs (heat range 7), to see if it would start and run more easily, now that I’d fixed a number of other things. Sure enough, the 7’s ran quite well and it only backfired a little at idle.
23. By now it was March 15 and I could actually drive the car. I took it for a 35 km drive and it was very strong. The AFR was as low as 11.0 at WOT at 4500-5000 rpm and the CHT’s never exceeded 300 F. Since it was only about 3 C outside that day, which would result in a lean condition, I think it’s safe to try some main jets that are one size smaller.
24. During this winter I also applied some Horseman’s cream to the ancient leather on the seats, checked the brake lines at the master cylinder to ensure they weren’t rusted, checked and tightened the front wheel bearings, and installed new rubber (OEM) front sway bar bushings, to replace the delrin type that Chuck had made, since one of those pieces had fallen out.
25. I have ordered fibreglass block-off plates for the engine, to eliminate the flow of air to the heat exchangers. This way the fan’s cooling entire effect will be directed to the cylinders. I have also ordered an aluminum floor pan to replace the original plywood piece which has started to delaminate.
26. In late March I got around to connecting the laptop to the Innovate system and discovered that the right side carburetors were significantly leaner at idle than the left side. After making adjustments to both sides, they are both idling at about 0.83-0.85 Lambda and the engine starts much more easily. I guess Chuck thought he could do it better by ear!?
27. I have modified the AFR display to show Lambda, since I will be blending race fuel and its stoichiometric value may not be 14.7. I plan to mix 110 octane unleaded with 91 octane Shell V-Power unleaded, to get a blend of approximately 100 octane.

It has been a busy winter – although it seemed to last forever – and I expect the car to run extremely well now that I have it tuned the way it should be. I’m looking forward to completing the break-in period, so I can use all of the engine’s power and get some CHT and AFR readings at that level. I intend to monitor CHT carefully and adjust the timing if necessary to ensure the temperature stays below 450 F. The first chance to do this will come April 16 at Calabogie. Needless to say, I’m looking forward to it.