Engine Block Testing

[Richard M.]

Re: Engine Block Testing

That was a great composition Duke, very informative, thanks.

[Steve M.]

Re: Engine Block Testing

X2, thanks for all the info guys. Hopefully, just one build on this project. Great to have a plan that everyone has checked out going in.

[Steve G.]

Re: Engine Block Testing

There have been many threads about "roller tip rocker arms" over the years, but since you're relatively new, I'll go over the arguments, again.

The tip of production rocker arms is ground in an arc that is very close to the arc of the rocker arm motion. For this reason the rocker tip "rolls" across the top of the valve stem. It does not slide, so a roller tip is worthless.


Roller tip rockers are a complete scam. They are heavy, which reduces valve train limiting speed, and many require a special thin wall socket to adjust the clearance/preload. The money is better spend massaging the heads and addressing known reliability issues with better parts like early (through '65) small block connecting rods and the early big block single spring/damper design that was prone to breakage.

For road engines I always recommend standard tension rings. Low tension rings are okay for racing engines that spend most of their life at WOT, but road engines spend most of their time at idle and light load, so there is often not enough gas pressure to force the rings against the cylinder walls to create a good seal. The result is often excess blowby and/or excess oil consumption.

As I've said many times before, road engines and racing engines are completely different animals, and installing "racing parts" in a road engine is often a recipe for poor performance, fuel economy, and excess oil consumption.

The more I thought about this roller tip rocker discussion the more curious I got. Not being privy to the original debate, ultimately, to what did you attribute the fact that worn out guides in stock engines are worn oblong, not concentricly, with the longest dimension in line with the rocker?

About 40 years ago I put together my first small block engine. Tried to do it right, line bored, 30 over, 12.5, ... but on a limited budget. The cam was a warmed over version of the 30-30 cam. Heads were done etc. Engine worked great and was everything I expected it to be. In under a year, however, the valve train was noisy(er) and I couldn't adjust it out. Turned out all the guides were worn out in my less than year old heads. Puzzled by this, I researched and read everything I could find (no internet!) about heads etc and found my answers in a piece by Crane Cams. I had used pushrods and rockers that would have been stock for that engine with it's original grocery getter cam. As you know, you can't make a high lift cam by increasing the lobe height, you have to reduce the base circle. And with a smaller base circle and stock length pushrods, in the closed position the angle of the rocker between pivot point and valve tip is now greater than 90 degrees. As the valve starts to open It is pushing sideways as well as down, which quite rapidly wore out the guides.

Where I was going with this long dry storey is that this experience, plus that of having seen the oval guides on stock worn out heads, left me with the impression that the roller tip would help reduce the friction that was causing the wear I've seen. Now, of course I now know you buy your pushrods as part of a calculated package, but things like shaving heads and decks change valve train geometry. I would have thought roller tips would provide some compensation for those small changes to the geometry.

I'm interested to hear what you learned.
Thanks,
Steve

[John H.]

Re: Engine Block Testing

I would in fact like to know exactly how a machine shop can save the stamp pad when it MUST be decked due to flaws like above, or to remove surface rust on old stored blocks etc.

How exactly is it done?

Rich -

The photo in your post #6 shows a typical Serdi-type surfacer, where the cutting head can only move in one axis (parallel to the crankshaft centerline); this is the most common type of surfacer found in most automotive machine shops. In order to deck a block and save the numbers, a machine with a second axis of motion for the cutting head (perpendicular to the crankshaft centerline) is required (see photo below); this type of dual-axis CNC machine is much more expensive than a Serdi surfacer, and not every machine shop has one. It allows running the cutting head from rear-to-front until the cut is just short of the "numbers", then switching to the perpendicular axis to create a straight perpendicular cut which will end up just forward of the front edge of the cylinder head, the same depth as the rest of the deck.


BlockDecking.jpg

[Duke W.]

Re: Engine Block Testing

The more I thought about this roller tip rocker discussion the more curious I got. Not being privy to the original debate, ultimately, to what did you attribute the fact that worn out guides in stock engines are worn oblong, not concentricly, with the longest dimension in line with the rocker?

About 40 years ago I put together my first small block engine. Tried to do it right, line bored, 30 over, 12.5, ... but on a limited budget. The cam was a warmed over version of the 30-30 cam. Heads were done etc. Engine worked great and was everything I expected it to be. In under a year, however, the valve train was noisy(er) and I couldn't adjust it out. Turned out all the guides were worn out in my less than year old heads. Puzzled by this, I researched and read everything I could find (no internet!) about heads etc and found my answers in a piece by Crane Cams. I had used pushrods and rockers that would have been stock for that engine with it's original grocery getter cam. As you know, you can't make a high lift cam by increasing the lobe height, you have to reduce the base circle. And with a smaller base circle and stock length pushrods, in the closed position the angle of the rocker between pivot point and valve tip is now greater than 90 degrees. As the valve starts to open It is pushing sideways as well as down, which quite rapidly wore out the guides.

Where I was going with this long dry storey is that this experience, plus that of having seen the oval guides on stock worn out heads, left me with the impression that the roller tip would help reduce the friction that was causing the wear I've seen. Now, of course I now know you buy your pushrods as part of a calculated package, but things like shaving heads and decks change valve train geometry. I would have thought roller tips would provide some compensation for those small changes to the geometry.

I'm interested to hear what you learned.
Thanks,
Steve

Valve guide wear is the result of many factors. There will always be some side force on any rocker arm design, but if the rocker arm contact point is at the centerline of the valve stem at about half lift, wear will usually be minimum. As built by GM, typical vintage V-8s were in this ballpark, so guide wear was not excessive, but it was usually the first wear issue to show up somewhere between 60 and 100K miles due to high oil consumption. Sometimes it was just a matter of new valve seals, but sometimes guide wear was excessive, which called for a head refresh long before bore wear became excessive.

High lift aftermarket cams, head and block milling will often upset this geometry, and most guys never check it.

There are also compatabilty issues between valve guide materials and stem material or surface coating, and different guide/stem materials have different oil flow requirements.

Lot's of variables.

I have my own explicable experince with valve guide wear on my Cosworth Vega. Due to high oil consumption during warranty GM agreed to replace the head, but I was aware that there were some issues with valve guide/valve seat concentcity as manufactured by GM.(I actually got Cal Wade on the phone to discuss the issue.) I asked the dealer for the head after it was removed and my measurements of the original cast iron guide inserts showed considerable taper and out of round.

I actually talked GM into paying for a local Cosworth expert to make new silicon-bronze valve guides and reseat the valves concentric with the new guides, and with new OE seals, which were not that great, oil consumption was tamed, but still fairly high.

Over the next couple of decades and 60K miles, including at least 5000 miles of race track hot lapping, oil consumption gradually increased and after I retired from racing got down to about 100 miles per quart, so I decided I had to do something. I thought it just needed new seals, but once I started removing valve springs to change seals it was clear that the valve guides were totally shot, so off with the head.

Now one has to consider the design, which is like other Cosworth heads of the era. In fact, the CV head is just a scaled up version of the DFV head to fit a Vega block. The valve lift is a modest .340" and the lobes engage a "shim over bucket" tappet that is about 1.4" diameter that is contained, along with the cams, in a single cast cam carrier with "integral" cam bearings. In other words, the cam carrier surface itself is the bearing. There are no bearing inserts. One would think that this large tappet snuggly clearanced in a bore would bear all side loads and the guides would last a long, long time, but no cigar, and I have never figured out why.

I was able to find some off-the-self nickel bronze guides, and also found a better seal. That was back in 2001, and I can't even measure oil consumption because of low mileage accumulation, but it seems to be nil.

...back to the current OEM design like modern Chevy V-8s. They use roller trunnion rocker arms with plain tips properly ground to "roll" across the valve stem and today's more precise machining (like deck height) allows closer to "perfect" average geometry than vintage engines. Chevrolet and other OEMs have spent a lot of money on better parts to reduce internal friction and wear, including roller trunnion rocker arms, which certainly cost a whole lot more to manufacture than the old precision stamped rockers, but these modern parts still have plain tips.

If roller tips are so great, why don't the OEMs use them on current engines? They certainly seem to have spared no expense in modern engines and adding roller tips to rocker arms would not be a major cost issue.

Guide wear is promoted by side force on the stem, but a properly designed rocker arm with a tip arc the same as the rocker arm arc, with the contact point at the center of the valve stem at half lift does not put a significant side load on the guide because the rocker arm tip "rolls" not "drags" across the stem. Side force is caused by force that is not always directed at the center of the valve stem, which is the case of any rocker arm design during part of the event. The Cosworth design should effectively eliminate any wear causing side force, but the CV guides certainly were much less durable than the original integral cast iron guides on my 327/340.

Duke

[Richard M.]

Re: Engine Block Testing

Rich -

The photo in your post #6 shows a typical Serdi-type surfacer, where the cutting head can only move in one axis (parallel to the crankshaft centerline); this is the most common type of surfacer found in most automotive machine shops. In order to deck a block and save the numbers, a machine with a second axis of motion for the cutting head (perpendicular to the crankshaft centerline) is required (see photo below); this type of dual-axis CNC machine is much more expensive than a Serdi surfacer, and not every machine shop has one. It allows running the cutting head from rear-to-front until the cut is just short of the "numbers", then switching to the perpendicular axis to create a straight perpendicular cut which will end up just forward of the front edge of the cylinder head, the same depth as the rest of the deck.

Thanks John, That's exactly what I wanted to understand. Great photo. Yes it certainly looks expensive.

[John H.]

Re: Engine Block Testing

If roller tips are so great, why don't the OEMs use them on current engines? They certainly seem to have spared no expense in modern engines and adding roller tips to rocker arms would not be a major cost issue.
Duke

Duke -

We came to the same conclusion as we developed the second and third generation Viper V-10's (and the subsequent current production 4th generation engine); we spent a lot of money on anti-friction coatings on the piston skirts, the "cam-in-cam" design for variable valve timing, husky forged rods with ARP bolts, Mahle-patented "beehive" valve springs, and many other details which brought us to 645hp from 505 cubic inches (naturally aspirated). Roller-tip rockers had zero effect in testing, so we didn't use them - rollers only on the trunnions, where it counts. We launched this valvetrain in 1994, and haven't changed it since.


ValveGear800.JPG

[Steve G.]

Re: Engine Block Testing

Duke -

We came to the same conclusion as we developed the second and third generation Viper V-10's (and the subsequent current production 4th generation engine); we spent a lot of money on anti-friction coatings on the piston skirts, the "cam-in-cam" design for variable valve timing, husky forged rods with ARP bolts, Mahle-patented "beehive" valve springs, and many other details which brought us to 645hp from 505 cubic inches (naturally aspirated). Roller-tip rockers had zero effect in testing, so we didn't use them - rollers only on the trunnions, where it counts. We launched this valvetrain in 1994, and haven't changed it since.


[ATTACH=CONFIG]57417[/ATTACH]

My prospective is not at all from the point of horsepower gain, but rather of valve guide longevity. Valve guides in these old engines often didn't last the lifespan of the engine, a lifespan which was already very short relative to today's engines.

The oe doesn't care much about engine live much beyond warranty. If they did lots of things would be done differently. Many of the things that we alter from the original design is not about improving horsepower but rather about reliability and durability.

The fact that guides wear in an oblong shape parallel to the rocker is pretty firm evidence that there is a lateral force from the rocker arm at play. If the guides in these old engines went 200K without a problem, you wouldn't give it a second thought. But they didn't. They mostly didn't make a 100K. So, while the arc on the tip was a reasonable effort to mitigate the forces at play, it was far from perfect.

One huge difference between the Dodge rockers, not just the ones in your photo, but most of them even back in the day, is that the pivot point is fixed. As I mentioned in an earlier post, guide life in a Chev engine was extremely short if a high lift cam was used with a standard length pushrod because the angle of a line drawn between the pivot of the rocker and the tip of the valve stem and a line drawn through the stem of the valve was increased. This caused an increase in side pressure on the valve and guide. I've personally been around the block with that and Crane Cams identified this as the cause.

What makes the gm design so poor and susceptible to this problem is that the nature of the adjustment moves the pivot point up or down. Not so with the Dodge design in your pic. The pivot point is fixed and assembly mandates a fixed height stem. No matter what you do to the head surface or block deck, that angle is always at design. Any change to the dimension from the camshaft to the rod end (block deck, head surface) of the rocker is taken up by the effectively increasing the push rod length with the lifter. On a sbc if you deck the block and shave the heads you simply end up tightening the adjuster further and increasing the angle between stem and rocker.

I suspect that the amount of tolerance of a stamped rocker is also higher than, say the nicely machined piece in the photo.

Steve

[Steve M.]

Re: Engine Block Testing

Machine shop called today and said one of the cylinders has a crack.

What considerations should be taken into account as far as having it sleeved. The block is still at std bore.

[Leif A.]

Re: Engine Block Testing

Machine shop called today and said one of the cylinders has a crack.

What considerations should be taken into account as far as having it sleeved. The block is still at std bore.

Steve,

Sorry to hear about the cylinder head. My question is why are you concerned about sleeving the engine block just because you have a cracked cylinder head?? One doesn't necessarily relate to the other.

[Donald A.]

Re: Engine Block Testing

Leif;

I think you misunderstand. Steve is talking about a cylinder in the block being cracked and having to sleeve it and or more. As with many machining operations I imagine there are plenty of wrong ways to do it.

[Steve G.]

Re: Engine Block Testing

Machine shop called today and said one of the cylinders has a crack.

What considerations should be taken into account as far as having it sleeved. The block is still at std bore.

Talk to your machine shop. The question now becomes what bore to go to. I wouldn't plan on sleeving the cracked cyl back to nominal if there is any wear at all in any of the other cyls, even if it falls within what is considered serviceable spec. If it is not .000 0n taper and out of round I would bore all the cyls to whatever the min it takes to clean them up and of course the sleeved one goes to that. In deciding what oversize to go to I would look at what sizes my choice of piston was available in. If it will clean up at .010 and the piston you want is available in a .010, that's where I'd go. If the piston I wanted started at .020, that's what i would go to.

Talk to them about the missing cap and ensure that they can deliver a serviceable block and how much it will cost.

Steve

[Leif A.]

Re: Engine Block Testing

Leif;

I think you misunderstand. Steve is talking about a cylinder in the block being cracked and having to sleeve it and or more. As with many machining operations I imagine there are plenty of wrong ways to do it.

Don,

After re-reading Steve's post, I believe you are absolutely correct...my bad.

[Steve G.]

Re: Engine Block Testing

Talk to your machine shop. The question now becomes what bore to go to. I wouldn't plan on sleeving the cracked cyl back to nominal if there is any wear at all in any of the other cyls, even if it falls within what is considered serviceable spec. If it is not .000 0n taper and out of round I would bore all the cyls to whatever the min it takes to clean them up and of course the sleeved one goes to that. In deciding what oversize to go to I would look at what sizes my choice of piston was available in. If it will clean up at .010 and the piston you want is available in a .010, that's where I'd go. If the piston I wanted started at .020, that's what i would go to.

Talk to them about the missing cap and ensure that they can deliver a serviceable block and how much it will cost.

Steve

Forgot to mention, you are also going to want to have the discussion about the block deck again with the machine shop. Typically, the cyl is bored to an interference fit with the od of the sleeve, but stops short of going all the way through to the bottom. This leaves a ledge to retain the sleeve at the bottom. The new sleeve is pressed in but stops short of flush with the deck when it contacts the ledge at the bottom. The protrusion of the sleeve at the top is removed in the decking process, leaving a perfect, almost invisible seam and a perfect sealing surface. The sleeve is then bored and finished and as long as the finished bored is larger than the original bore size the small portion of the original bore left at the bottom is flushed out with the sleeve and becomes almost undetectable. There is some controversy about the need to finish the sleeve to a dimension smaller than the original bore to dress out the retaining ledge. I personally don't think it's necessary, but that's just my opinion. And it's a non issue if the other cyls need refinishing.

So, the condensed version of this is, can they dress the top of the sleeve to deck height and retain the stamp pad?

[Patrick B.]

Re: Engine Block Testing

Forgot to mention, you are also going to want to have the discussion about the block deck again with the machine shop. Typically, the cyl is bored to an interference fit with the od of the sleeve, but stops short of going all the way through to the bottom. This leaves a ledge to retain the sleeve at the bottom. The new sleeve is pressed in but stops short of flush with the deck when it contacts the ledge at the bottom. The protrusion of the sleeve at the top is removed in the decking process, leaving a perfect, almost invisible seam and a perfect sealing surface. The sleeve is then bored and finished and as long as the finished bored is larger than the original bore size the small portion of the original bore left at the bottom is flushed out with the sleeve and becomes almost undetectable. There is some controversy about the need to finish the sleeve to a dimension smaller than the original bore to dress out the retaining ledge. I personally don't think it's necessary, but that's just my opinion. And it's a non issue if the other cyls need refinishing.

So, the condensed version of this is, can they dress the top of the sleeve to deck height and retain the stamp pad?

I have had several blocks where the top of a sleeve was dressed to deck height without the need to deck the block and mess up the pad. Perhaps they were finished with a conventional mill bit rather than the big fly cutter used for decking.