supplemental coverage

Even with the entry that I just posted today, I am probably still at least two or three entries behind on this blog.  It takes at least a couple hours to write an entry of any considerable length, and I often just don't feel like sitting down and doing it.

However, in the interest of providing more information about what's going on, and in a format that will be incredibly easy for me to update, I have just set up an Instagram account which will allow me to quickly post photos.  Most of these will probably be related to the Matilda project, and will probably serve as "sneak peaks" at stuff that will be covered in future blog entries.  Some of it may just be stuff that I think it looks cool.  Maybe that will end up being most of it, actually, I don't know.  We'll see.

The good news is, you don't actually need an Instagram account to see the photos (...I think...).  You should be able to see them just by going to this URL:  http://www.instagram.com/golikehellmachine

OK, then what?

Warning:  this entry is pretty dry, bordering on technical.  I tried not to go into too much fundamental explanation, figuring that anyone who doesn't already have a pretty good handle on engine operation will probably not be interested, anyway.

OK, so, the oil pan was eventually painted and mounted, and then ... what?

Well, I put the engine back on the frame, and I was getting ready to bolt the bellhousing and transmission up when I remembered that I'd never checked pushrod length.  But, to check pushrod length, I'd need to be able to turn the engine, and my transmission came with all kinds of warnings about what not to do before it's been spun by the engine in neutral with oil in it, to the point where I decided I didn't even want to turn it by hand at this time, if I didn't have to.  And if it was bolted to the engine, then the engine would be turning the input shaft as I rotated the engine.  So, for the time being, I supported the motor plate (which forms the front face of the bellhousing) with a jackstand until I could finish checking pushrod length.

Next, all I had to do was check pushrod length.  I thought I remembered how to do that, but decided first I'd read up and make sure that what I thought I remembered was correct.  Boy, was that ever a mistake.

Once I started looking around online, things only became less and less clear.  As with most things online, check three message forums, get 37,462,954 different answers.  The method I thought I remembered was to check the travel of the rocker tip across the valve stem tip, and to center that travel on the valve stem tip.  The idea is that you don't want the rocker pushing on the valve stem tip off center, because that will cause the valve to rock in the valve guide, which will prematurely wear out the valve guide.  The thought is that if you don't set pushrod length correctly, the valvetrain will wear out much faster.  And if you do set it correctly, you'll have an invincible engine that runs forever and ever, and ever, and ever, and ever.  And ever.  And--in some cases--ever.  And I want that.  So I decided I'd better be very careful about getting my pushrod length correct.

You can check the travel by coloring the valve stem tip with a marker, then turning the engine by hand through two crank revolutions (one cam revolution), and then checking where the marker ink has been wiped off the valve stem tip.

But, if you're using hydraulic lifters, like I am, then you need to do something to keep the lifters from collapsing while you turn the engine.  If the lifters are allowed to collapse, then th rocker will not move through its full range of motion and you won't get an accurate representation of where the rocker tip travel is positioned on the valve stem tip.  So, one method is to use solid lifters just for checking pushrod length, and then install the hydraulic lifters.  Another method is to use "checking springs" in place of your valve springs.  Checking springs are very lightweight springs which you can compress by hand, and which will not collapse the hydraulic lifters ... as much as the valve springs would.

Here's a photo showing the standard valve springs (cylinder 8, the two springs on the left) and the checking springs (cylinder 6, the two springs on the right):

Before this project, I had two different styles of valve spring compressor.  One style can only be used with the cylinder head off of the engine.  It's shaped like a big "C," and it reaches around the cylinder head to push up on the head of the valve while it pushes down on the valve spring retainer on the other side of the head.  That wasn't an option here, as the heads are already on the engine.  I had another style which is made to grab the sides of the valve spring and pull up while it simultaneously pushes down on the valve spring retainer.  I thought that would work, but then realized that since these heads have double valve springs on each valve, an inner and an outer, it wasn't going to work because it couldn't grab the inner spring.  So, I ended up ordering a third style of valve spring compressor.  This one threads down on to the rocker stud and has a fork that presses down on the valve spring retainer.  I pulled the spark plug from whichever cylinder I was working on and stuffed a length of windshield washer hose into the cylinder to keep the valve from dropping into the cylinder.  The really nice thing about this tool is that the lever for compressing the valve spring will lock into position at full travel, so you can have both hands free for fishing the keepers out.

Anyway, here's a view of the tops of the cylinder 6 valve stem tips, with black stripes drawn on them with a Sharpie:

And here's the results of turning the engine through one complete cycle (two crank revolutions, one cam revolution), with the standard length pushrods:

These look pretty good, but the mark on the exhaust valve (left) looks like it's shifted a little bit down, toward the outboard side of the engine, and the mark on the intake valve (right) looks like it's shifted a little bit up, toward the inboard side of the engine.  This would indicate that the pushrod on the exhaust valve is a little long, and the pushrod on the intake valve is a little short.

The critical piece of equipment for checking pushrod length is a couple adjustable pushrods.  You can't run the engine with these, but you can use them to try different pushrod lengths while checking.  There are different styles, but the ones I got have one piece threaded into the other, and they are a certain specified length when screwed all the way in.  Then each turn that you unscrew them adds 0.050" to that length.  So you change the length and check the result, change the length and check the result, and repeat until it looks good, and then you count the turns back to minimum length, do a little math, and that's the length you need.

Here's a picture of the marks with both pushrods at a pretty good length: 

But wait ... we're not even CLOSE to finished.

Next a million things happened to complicate everything.

First off, in spite of the fact that Comp Cams states in a (pretty disappointing) video on YouTube that their checking springs are weak enough that they won't compress hydraulic lifters, it turns out that that is not true.  I eventually discovered that I was not reaching full valve lift because the lifters were still collapsing, even with the checking springs installed.  That means that all the stuff I did above was inaccurate.

But, like I said, there are several theories on what's the correct way to determine "optimal" valvetrain geometry.  The reason I discovered that I wasn't reaching full lift was because I wanted to try one of the alternate methods, which was to set pushrod length so that the rocker arm is at a 90-degree angle to the valve stem at half lift (or, an alternate theory says two-thirds lift, to shift the optimal geometry towards full lift, where spring forces are at maximum).  I will also mention now that another theory says that you want to minimize the width of the rocker's travel across the valve stem tip, regardless of where this puts the rocker on the valve stem tip.  If you think about it (and I did think about it ... waaaay more than it was probably worth), this "minimum width" theory should give the same result as the 90-degree-at-half-lift theory.  They are just two different ways of measuring the same result.

Anyway, to check where your rocker is at half lift, you have to know when you're at half lift.  To check when I was at half lift, I used a dial indicator.  The dial indicator showed, however, that I wasn't getting anywhere near full lift, which was how I realized that the checking springs were still strong enough to collapse the lifters.

Here is a photo of the engine set up with the dial indicator riding on the valve spring retainer for the cylinder 1 exhaust valve:

And here, for the record, is where the rocker travel ends up if I tried to minimize its width, and/or put the rocker at a 90-degree angle to the valve stem at half lift:

You can see that the mark is way off center, shifted to the outboard side of the engine, indicating a short pushrod (according to the other theory).  In fact, the pushrod was so short that the rocker was rubbing on the lash adjusting nut when the lifter was on the base circle.  That's the kind of thing that tells you ... "something's not right here."  Now, it's possible that by choosing a different rocker, or changing some other variable, I might have been able to shift that mark toward the center of the valve stem tip while still retaining a minimal width of travel and holding my 90-degrees-at-half-lift target.  But all these parts are supposed to be correct for this engine, and I didn't feel that it was likely that I should have to do all that.  I decided that it was more likely that these methods were incorrect, and that they were being championed by people who were either misinformed, misguided, or possibly familiar with this method from working on some other type of engine.

As I was questioning these other methods, I went back and very carefully reviewed Comp Cams' instructions for optimizing pushrod length.  Their instructions described the rocker arm moving from the inboard side of the valve stem tip towards the outboard side of the valve stem tip at full lift, and then back towards the inboard side.  If the rocker were at a 90-degree angle to the valve stem tip, with a minimized width of travel across the valve stem tip, then it would have to reach its outboard-most point of travel at half lift, moving back to its inboard-most point at full lift, then back to its outboard-most point, and back to its inboard-most point when the lifter was back on the base circle.  This is not what they described, so I threw out the 90-degree/minimum-travel theories and returned to the centered-travel theory.

As an aside, one of the people I'm working with from the client on my current project at work used to be involved in Ford's NASCAR program.  I asked him one day if he knew how the Ford NASCAR guys defined optimum valvetrain geometry.  He said that rocker tip travel was supposed to be centered on the valve stem tip, which confirmed the theories that I was favoring by this point.  I asked him, "That's how the Ford guys do it?"  He said, "I think so, yeah."  I faked a look of concern, and then I asked him, "Well, did you ever happen to find out how the Chevy guys do it?  You know, if I wanted to do it the right way?"  He rolled his eyes and threw up his hands and said, "Well, I could call somebody, if you really want to know...!!"  I said, "No, no, no, I wouldn't want you to interrupt them, they're doing important work."

Anyway, now I'd finally decided what method I was going to use, it was just a question of crunching through it.  Working through the cylinders, I put checking springs on each one, put a dial indicator on each valve, and turned the engine to where it stopped lifting the valve.  That should be the full lift position for the camshaft, and then I could just push the valve spring retainer down by hand until the valve reached full lift.  You can actually feel the spot where the lifter plunger spring is no longer helping you to open the valve, representing full lift.  By this method I could check the position of the rocker tip on the valve stem at full lift.

I found some small variations from cylinder to cylinder, but averaged them all out and ended up with measurements for intake and exhaust pushrods slightly shorter than the ones recommended by Comp.  This makes sense, as I believe the engine block has been decked, which would shift the cylinder head down and require slightly shorter pushrods.  So I sent the measurements to Comp and they sent me the corresponding pushrods, which are now in the engine.

All in all, I still can't really be sure if I "got it right."  Did I use the right method?  Like so many debates on the internet, nobody has hard facts to substantiate their argument for why their method is "right."  Did I get the measurements right?  I did my best, but can't be sure.  But I took all the information available, tried some things, tried to decide what made sense, made a decision on how to proceed, and then considered the result, to try to decide if it made sense.  I think it does.

All this deliberation probably added at least a month, maybe two, to the project.  Was it worth it?  Well ... probably not, really.  The more I considered all these ideas, the more I slowly came to feel that it probably doesn't really particularly matter.  I suspect that the importance placed on valvetrain geometry is probably a carryover from the days of stamped steel rockers.  The tip of a stamped steel rocker had to be carefully contoured so that its contact point with the valve stem tip would always be oriented to press straight down on the valve stem tip, regardless what angle the rocker was at as it traveled through its full range of motion.  That contour on the rocker tip can only be wide enough to be effective for a certain window of rocker arm travel, i.e. a certain window of pushrod lengths.  Outside of that window, the rocker arm tip would be a flat surface pushing on the valve stem tip's flat surface at an angle, which could push sideways on the valve stem tip and apply a torque to the valve, wearing the valve guides.  Nowadays, though, roller rockers are so popular, they virtually render stamped steel rockers irrelevant.  Roller rockers have a small roller built into their tip, which ensures that the force transmitted from the rocker tip to the valve stem can only be directly in line with the valve stem.  I suspect this design feature makes valvetrain geometry much less critical.  Just like everyone else arguing these points on the internet, however, I have no hard evidence to support my claims, which is why I still went through a month or two of trying to determine the optimal pushrod length for my engine.

Anyway, at this point pushrods are done, I feel pretty good about what's in there, and I'm moving on.