It's funny how the more tired the clich, the more likely there's a kernel of truth to it. The case in point here is that timeworn favorite: If you want to finish first, first you have to finish.

What good is insane horsepower if won't hold up for the length of a race? Likewise, Honda-like dependability isn't much fun, either, if it's so low on power you are a sitting duck on the track. To get the power you need to be competitive, an engine builder always rides that ragged edge of dependability. Parts are pushed to their breaking point; the trick is to catch failures before they happen.

"I talked about this just the other day with one of the leading rod manufacturers," says Keith Dorton, owner of Automotive Specialists and a member of Circle Track's Technical Council. "We were talking about some new pieces, and I told him, 'If you want to test something and see if it will tear up, give it to a racer.' It doesn't matter what it is, if a racer gets hold of it, he's going to push it until it breaks.

"That's not necessarily a bad thing, it's just the nature of racing," he continues. "What that means, though, is if you can spot potential failures before they happen, you can save yourself a lot of money and headaches."

Observation is key Because Dorton is an engine builder, this story will concentrate on failure prevention of engine parts, but Dorton says the following principles apply to every part on a race car. Basically, it all boils down to observation and proper handling methods to avoid weakening parts.

Often, engine parts that are about to break will show signs of fatigue-usually cracks. The best time to spot them is at engine teardown. "I use a visual inspection at teardown because that's the time it's usually easiest to spot. It's harder to see cracks after each part has been cleaned and polished up. If you wipe off a part that has a crack during teardown, some of that oil is going to stay in the crack, which really makes it show up," Dorton explains.

"For a good visual inspection, all you really need is a good light and some sort of magnification. I use a lamp that's got the magnification built in. Get to know where the failure areas are so you will know where to look."

Engine areas to look out for are the pistons (Dorton recommends checking for cracks, discoloration, and the ring lands for excessive or unusual wear patterns.), the snout of the crank (especially on Fords), and the head where it is bolted to the block. On Chevys, another area prone to cracking is between the exhaust valves on the No. 2 and 3 combustion chambers.

Changes Can Hurt Be aware that changes in one area of the engine can potentially be the cause of failures in another. "Pistons are a prime example," Dorton says. "A few years ago we had a situation where we had been a few seasons with a certain class of engine and never had any trouble with the pistons. But we were seeing more and more piston failures. Usually, our inspection process caught them before they failed so there were no catastrophic failures, but the pistons just weren't living as long as they used to.

"There had been no change in the pistons, but what we didn't realize was having an effect was gradual horsepower gains over that period of time. With the horsepower gains racers had also gradually increased rpm until the combination was just more than the pistons could handle. We had gone beyond the design limits of that piston, and we started having problems.

"So we went back to the manufacturer with some examples of what we were seeing and said, 'Look, here are the weak areas.' Luckily, the fix didn't require a redesign; it was as simple as adding another 0.020 or 0.030 inch of material in some of the weak areas and the problem went away."

When rebuilding an engine he had never had in the shop before, Dorton noticed a notch on the beam of each connecting rod just outside the oiling hole. It turns out the engine builder had bushed the small end of the rods, and when he drilled the oil holes the drill chuck cut a small notch on every connecting rod. Sharp edges create stress risers on parts, and Dorton guarantees that, left alone, these rods will fail. The solution was simple: He simply polished the notch smooth so all the marks ran with the length of the rod, and then shot-peened it. The moral of the story is to be extremely careful how you handle parts so as not to introduce a weakness by accident. Dorton even goes so far as to handle every valve separately (especially in the parts washer). Even clinking the edges of two valves together can potentially introduce a stress riser and invite trouble.