B3 Racings's Indy EZ's & T&D rocker arm geometry correction

I posted on here last fall regarding a valve train instability issue that Dwayne Porter deduced was the result of valve float due to too low of spring pressure resulting from incorrect valve spring height and possibly bad valve springs while running a hydraulic roller cam.

It almost cost me a motor, but I caught the carnage beginning when I did a routine inspection and pulled the valve cover. I had blown the needle bearing cages out of all the exhaust rocker arms, bent several pushrods slightly and could see the cage material laying in my heads. I feel I was really lucky, as the car had just taken a first place in our annual Mopar Alley car show, and I had blipped it up to about 140 mph(on the track) on my way home. It was still running great, so I thought.

Under Dwayne's review of the cause of the issue, he found that this problem was already apparent on the initial dyno pull sheets, as indicated by the sudden drop in power at 5400 rpm. The motor was making 684 hp just before that drop.

I had Dwayne go completely through the heads and replace ALL the hardware and set them up for use with a solid roller. I went with the Isky red Zone bushed lifters for my street engine. Dwayne is a great man to do business with, and a true visionary in his field.

While the heads were off, Dwayne helped me spec out some custom Ross replacement pistons to reduce my compression enough to switch from race gas to pump gas. The compression drop was from 12.2 to 1 down to 10.2 to 1. The pistons came in around 514 grams. This motor has a very low reciprocating weight, and should love to rev, so I wanted to make sure I never had valve train instability problems again. I used top shelf components and took the extra step of correcting the rocker arm geometry.

I had my 1.6 to 1 T&D roller rockers rebuilt at T&D and put them on the heads. The geometry seemed way off with the roller tip reaching well above the centerline of the rocker shaft and the adjuster screw at a pretty step angle in relation to the pushrods. I still have the rebuilt set if anyone wants a deal on them.

I contacted Mike at B3 Racing and followed his instructions to get the measurements he needed to make a correction kit. As it turns out he was working on a similar project at his shop. He said in my case the geometry was so far off from ideal, that I would need to have custom rockers made by T&D. He figured out all the measurements and I had my parts a few weeks later. Mike is another great guy to do business with, and I highly recommend both men if you are locking for this type of work.

I know a lot of people on here are skeptical of this type of geometry correction, but the results speak for themselves. All you have to do to see if correcting your rocker arm geometry is worthwhile, is to look at my before and after pictures. It's such a dramatic difference, in my mind a guy would be crazy not to make this correction.

I consulted with Manton Pushrods shop foreman and upsized my pushrods to their 7/16" Stage 5 pushrods. Here is another business that provides exceptional customer service. Dwayne had increased the pushrod clearance on my heads when he had them apart and the thicker pushrods have plenty of clearance now.

I'll let the pictures of Mike's geometry correction solution do the rest of the talking on the subject of, "Is it worth taking this extra step".

Thank you for that. Food for thought. Can you tell us more about the combination
In terms of design lift, ratio and so on?

I was taught by the old Mopar drag engineer group at their drag race seminars that wedge motor single shaft rocker arms should start with the contact pattern on the valve stems to start on the inner third of the valve stem and then sweep outward across the center of the stem to the outer third at max lift, does yours do that now?
I have a set of 440-1 with a set of T&D single shaft 1.7 ratio that I haven't used yet, I have another set of them on a motor with a set of Jesel 1.55 ratio paired shaft system and the roller tips don't move much at all off the center of the valve stem at all
The other motors I've built and raced with 440-1 with Harland Sharp 1.65 ratio single shaft systems did like Mopar taught me
I'm glad you caught those problems before it hurt the motor or you
I have broke parts on the track and thought I was going to wad up one of the race cars , thanks to God that didn't happen

Can you post a picture of the sweep pattern. The width is what's important. It's more important than having the pattern centered.

Glad to see you're finally getting it back together!!

It better not, or Mike botched his kit!

Two current schools of thought on rocker geometry:

1. Have the rocker arm perpendicular to the valve tip at 50% lift, which means the rocker tip starts slightly inboard of the center of the valve tip, sweeps out across the center at mid-lift, then sweeps back slightly inboard again at max lift. This approach yields a narrower overall sweep pattern and -- if components are designed properly -- will have this sweep pattern centered over the valve tip. I'm pretty sure this is the approach Mike at B3RE takes w/ his kits. Although T&D's bolt-on kits don't achieve this, this is the "ideal" geometry T&D describes in their catalog.



2. What I'll call "Jesel Geometry" (since I've seen them use this term themselves) which has the rocker arm perpendicular to the valve tip at around 2/3 max lift. Their explanation for this approach is that it reduces the amount of sweep when the valve train is under the highest spring loads, although it results in a wider sweep pattern than what I described in #1, and also puts the starting point of the sweep when the valve is closed more inboard of the valve centerline.



The image directly above shows the sweeps that result from both the Jesel "Low Pivot" approach and "Half-Lift" approach that I described in #1.

Not quite: the contact pattern on the valve stems starts on the inner third of the valve stem and then sweep outward across the center of the stem to the outer third at HALF lift, then back to its original position at FULL lift; it will be centered on the stem tip at (nearly) 25%, and again at 75% lift.
The sweep distance is greatly affected by the ratio of the long (valve-side) lever divided by the net lift at the valve: short rocker (LA) + high (.700") lift = long sweep, long rocker (hemi exhaust) + short lift (.500") = short sweep, etc. This is not an adjustment or alignment, and the cure is a different rocker and lots of work.

Mid-lift is only available if you're using Jim Miller's rockers, it's a patented feature and does not exist unless you make it yourself. Setting the valve-side geometry to 1/2 up, 1/2 down @ 90 degrees just moves the error to the pushrod side. This is frequently an improvement, but that's all you can do - bump the discrepancy back & forth across the shaft axis.

A thank you to all whose expertise is helping to clarify a complex (to me)
interaction on valve train geometry.

Many of my gaps on this subject are becoming more clear due to everyone's willingness to share their knowledge.

A complex subject and still many opinions, but if we keep at it in a civil manner and with an open mind, this too will resolve into a body of info that will benefit enthusiasts.

Thank you gentlemen!!

Joe

I'd say the Jesel approach has been proven on countless high speed race engines in NASCAR, NHRA, etc.

The Jesel low pivot approach is also what GM Engineering recommended years ago in their GM Performance books. GM recommended setting the 90 degree point at 1/3 lift.

I did the calculations years ago to figure out the "perfect" height that would minimize the product of travel and load and it was close to the 1/3 height that GM Engineering recommended. There has been a lot of discussion lately about static loading vs. dynamic loading since once the engine is running at revs there are a lot of other things moving around.

Having said all of that I prefer to run the rockerarms the way the mfg designs them. T&D knows what they are doing and so does Jesel so I don't see any reason to mess with it. Lots of T&D and Jesel equipped engines making lots of power and staying together for lots of rounds so I don't see any reason to re-design what already works.

John, have you measured the net lift at the valve, after lash, with full spring load, with the new set-up?

Well, what they hey I'll chime in too. For me, us, we, we like to start with the shortest valve that can exist in the range of .05, or sometimes less before coil bind for the given lift. Idea being keeping all things as short as possible. Rocker shafts close to their anchors in the head, and less deflection, and weight. Personally I like to start the motion from the back of center of the valve, shooting for the shortest sweep available, even if it's off center. That's just me, so do what works for ya.

No one ever talks about the other half of the rocker arm geometry, as if it does not exist. Maybe it's un-important, although there seems to be no shortage of issue/failures on that side.

Although I'll probably give Mike a call the next time I do some rockers, I'm kinda with Andy on this one. For at least 40 years engineers in both auto manufacturing and aftermarket have been messing specifically with this topic to achieve different results for specific reasons. At the end of the day, my junk would never know the difference.

What he said...

I have a customer and friend who teaches the H.P. course at one of the SO CA ITT or UTI auto tech. schools (CRS ) who broke a valve in one of the motors I built for his 1968 dart BB drag car, he is putting it back together himself now and calls me when ever he has a question. He has found thing I missed , crank snout had .0015 run out at the timing gear and he saw the timing change tension vary from that difference
He is a book reader and cautious on some of the clearances, (he worked over twenty years in one of the SO CA Jaguar dealership before starting teaching, he has had the ASE master tech rating since day one) he was concerned with .060 intake valve to piston clearance and .085 clearance on the exhaust side with checking springs using a dial indicator on the retainer aligned with the valve stem angles front to rear and straight up and down. He clayed it also and saw some variance between the two ways so he called. I ask him to put the roller cam springs in ands redo the clay, it picked up over .020 more clearance
He is happy and worry free now
When in doubt, check it out

No way would I be confident with the shaft raised that far off the original pedestal and off center. If you want to correct it that much machine the mounts on the head flat and machine new mounting stands.

Forgive me for the lack of multi quotes, but I just want to clear up a few things.

Cab, the factory engineers were talking about non roller type rockers. They are supposed to be set up the way you described, but a roller will have a bunch of velocity at full lift when set up that way, and be very unstable. The rocker is supposed to transmit the cams lobe information as accurately as possible everywhere, not just full lift. It can't do that if the valvetrain is that far off.

Brad,

Once again, you are spot on. The reason the T&D rockers (or Jesel)don't achieve perfect geometry with the single shaft, is because they can't. They make a compromised rocker, for the guys who want to just bolt something to the existing stands, just like everyone else. I've talked with their engineer multiple times, and he admits that is true. But, they will tell you, "we have a paired system that has correct geometry". Yes, but you will have to use pushrod or spray bar oiling, likely have to bush the lifter bores with the pushrod oiling, cut off the stands on the heads, and class racers like Stock Eliminator aren't allowed to do that.

The Jesel set up may center the roller at the highest spring pressure, but does not decelerate the valve as much at full lift, or accurately transmit the cam lobe to the valve. So, what is more important, loading on the valve tip, or stability and accuracy?

We've talked about this before. Jim Miller's patent does not go on in perpetuity. It expired in 1994, and anyone who wishes to design a rocker with that concept, is free to do so. I have expanded on that concept, and that is what John (OP) has posted here. You are correct, though, that you won't find an off the shelf, single shaft rocker with the proper design.

Nothing new here - this method of raising the shafts up using an offset cam insert was tried more than thirty + years ago - you will eventually find that you will either pull or snap the studs out of the stands or push the whole shaft over - the more spring pressure the more the design flaw will begin to show. The solution thirty years ago was to machine the pads and make new blocks that were located and pinned to the head - yes it was pricey but it worked with lifts approaching .900. On a mild - low lift motor this method probably works fine - its an age old argument with a single shaft valve train as to rocker path - but there are many engines running T&D, Jesel, HS that are performing just fine. Its just a matter of personal choice

I learned most of what I know about rocker geometry and valvetrains from a Gibbs camshaft and valvetrain specialist in their engine department. They might use Jesel rocker arms, but they don't set them up they way Jesel recommends. In fact, many teams now make their own valvetrain parts in house.

I have a Late Model dirt track motor in the shop right now, that was built elsewhere. It ate an exhaust valve, and was using Jesel paired rockers set up the way Jesel recommends. The crew chief said the only failures they have experienced are all valvetrain related. With a flat tappet cam rule, these guys can't keep loading the motor up with spring pressure to mask the problem. It has to be right, and obviously it isn't.

I spent about a half hour on the phone with Larry Torres two weeks ago, discussing rocker geometry. He is well aware of the problems with geometry on Mopar single shaft systems, and the compromises made to give the market what they ask for, a rocker that can be "just bolted on". It doesn't make it right, but no one wants to believe that they will have to spend big bucks for rockers, and then have to modify things to set them up properly. Especially when they can read a book that will give them bad information, and make them feel better about their pain when they have problems. An opioid for the Mopar masses.

On the contrary, I talk about it all the time. But, it is the side that the geometry is designed into the rocker itself, which is why the OP is showing the custom B3RE/T&D rockers, so with off the shelf stuff, you correct what you can. The valve side being off will tear stuff up quicker than the pushrod side, so it is the higher priority.

Over five years in production, and not a single failure to date. Some of those motors are in the low nines, and others have thousands of street miles. How many manufacturers have that kind of record?

Thirty plus years ago, roller rockers for a Mopar were not all that prevalent, and .900" lift was unheard of in anything less than Top Fuel. As far as today, with today's technology and parts availability, who in their right mind would use a single shaft rocker system with 900" lift. Seriously!

Again, everything has its proper application, and I've done the blocks when necessary, or desired by the customer, and ZERO failures to date.

On a side note, I was supposed to have a set of the custom rockers at Carlisle this week, to compliment the Trick Flow 270's I will have on display. Unfortunately, T&D will not have them here in time, even though I ordered with plenty of time to spare. It would have been a good opportunity for show goers to check them out first hand.

I'm apparently doing it wrong again...

The amount of anti-mid lift, in which the 50% lift does not occur at 90 degrees, is frequently used by OEM to improve the pallet (non roller tip)-to-stem wear properties. Heldt wrote a formula to calculate (some type of) anti-scuff by changing the pivot height, but it's over my head!
How much of the "error" (non 90 degree) you want on the pushrod side will vary somewhat between engines depending on other problems such as the horrible LA 14 degree difference between the 59 degree tappet, pushrod angle and 15 degree stem angle.

I'm working on a geo article now with a single shaft, but valves at different angles (no, it's not a poly - guess what?), so moving the shaft to improve the intake requires correction back the other way on the exhaust, blah.

The Gen-2 426 has its own problem: the rocker stands (both OEM and many aftermarket) are a single piece, which means any height adjustment (mill or shim) affects both valves simultaneously to the same distance, causing much hair to be pulled out!

You are absolutely correct. Different angles, with a common shaft, causes a lot of grief. The Gen 2 Hemi would need new stands made every time a change is made, to get everything right, hence the hair loss. The higher the stands, the closer the shafts would have to get together. That's why Barton's Hemi rockers are all individually mounted. They are pricey, but allow adjustability for each individual valve.

The poly heads would be even more of a headache, unless the stem heights were all set at the exact spec needed, on both intake and exhaust, to get the single shaft at the correct height for both. Then the rockers would have to be designed with the correct fulcrum length for the finished cylinder head. Change the valve length or tip height, do it all over again. Change the valve lift, do it all over again.....etc. Brace yourself, I think I see a bald spot.

You read through this thread and you can see why Chrysler engines get virtually nothing in aftermarket parts, and what they do get is compromised junk.

Chrysler guys want a bolt on, good enough 30 years ago part. Then they talk like GM engines are snap together. Just bolt junk on and they make power.

None of that is true.

What is true in the number of Chrysler people who bolt on junk, then tell everyone they do it so it must be correct.

We get exactly what we deserve.

Let me mention that, in addition to those unfortunate enough to experience actual failure, seizure, excess wear, etc. in valve gear, there are also less-than-perfect installations that may show no immediate signs of error but in which minor tweaks will provide more power, more stable high RPM, etc.
Whenever you have it apart, take a moment to analyze anything that "doesn't look right" (scuff marks, dissimilar wear patterns on stem tips, burned/blue pushrod balls and cups, fasteners that continue to loosen, valve adjustment that changes over time, etc.). Any anomaly is cause for inspection and discussion.
These archives contain a huge volume of experience, real world data, history of experiments, etc. well worth your time to research.

Dont worry Al - I been doing it wrong for almost 40 years also - and yes roller rockers were available - common and used over 30 years ago and a 750 lift Crane R296 with custom HS 1.7 (yes they made them) got you close to an .850 lift - the idea was to keep the valve open at peak flow as long as possible - this was using iron heads with longer valves and 2' installed height - those were the days of innovation!

Roller-tip rocker arms: 1928 Indian 101.

https://en.m.wikipedia.org/wiki/Desmodromic_valve

Desmodromic valve control 1896. Interesting read. There is not much that has not been around for decades or even centuries.

I'm pretty sure Bentleys had them before that........along with ohc and 4 valves per cylinder.

Question, how does the cross-section (looking down the shaft line) of a rocker change with increased ratios?

It would seem as it only can be done by shortening the pushrod side, or is there another method?

Thanks, Joe

Oh I think I'm ok. This is definitely not anything new and a lot of ways to skin a cat

did you ever see a 20's rols royce engine - everything had roller bearings even the mains - was a piece of machine artwork - unbelievable craftsmanship went into those engines

I tried to get some clear pictures of the rockers with the valve closed, open half way and all the way; but they did not come out well. I'll try again this weekend.

I also tried to set-up my dial indicator to measure the actual valve travel, but my set-up kept moving and messing up my readings. I'll see if my friend Roger a.k.a. Sxrxrnr, will bring his dial indicator set-up over this weekend to measure it accurately.

The cam specs for the new solid roller selected by Dwayne for my application are on this link at Comp Cams.

http://www.compcams.com/Company/CC/cam-specs/Details.aspx?csid=743&sb=2

I like running stock valve covers, and it looks as if I will still be able to with a bit more massaging of the internal baffles. I still get a kick out of telling people at car shows its basically all stock with headers, and then watch their faces when I fire it up with the electric cutouts open.

Roller bearing crankshafts, another of those ideas that sound great, until:
1. they mention that the crank comes apart rather easily, sometimes in the middle of a trip
2. it increases the cost by a factor of 10
3. load capacity of roller vs. plain bearing: 6" mains and 4" rods would be a good start, it only worked because the engine turned low RPM and developed no power. Ask Enzo about roller cranks, but be walking away so he can't hit you
4. brain surgeon + many fixtures for assembly or service
5. rollers do not contain oil pressure, which means individual feeds from the pump to every journal, the crankcase looks like a subway line
6. very noisy, very difficult to diagnose problems

W. O. Bentley was respected as an engineer for his money.

Talk about high tech in the most unusual places; bought a Honda string trimmer 15-odd years ago, 25cc 4 cycle engine, OHC!!

Joe

shortening the pushrod side: easiest and most common
is there another method: completely new rocker, keep the stock lever total length to retain vertical alignment with the stem and pushrod, but move the shaft toward the pushrods

E.g. my current project:
Total effective rocker arm length, adjuster to pallet: 3.40625"
Pushrod lever 1.375"
Valve lever 2.03125"
Ratio 1.477:1

New rocker with .158" offset center barrel:
Total effective length: 3.40625" (same)
Pushrod lever 1.2165"
Valve lever 2.1900"
Ratio 1.8:1

Might not be worth your time, John. Apparently, these rockers have been made before, which is odd, because I paid a custom charge to have T&Ds engineer draw up a NEW rocker profile to MY specs. Regardless, it's nothing new, and your free information has no value here. Go out and enjoy your car, because these guys already have it all figured out. Too bad they didn't tell us a long time ago.

I've ran a similar grind solid roller in my old pump gas stroker 400 motor, it had 260@.050 intake side with .420 lobe lift, 266@.050 with .409 lobe lift ground on a 108 LSA installed at 105 and 106 ILC. That cam worked very well, especially when I changed the rocker arm ratio from 1.5 to 1.65
As far as stock covers and the internal baffles, the rocker arms you use will influence the results to get them to clear
I've had customers who insisted on running stock covers that ended up drilling large enough holes in the baffles to clear the adjusters without removing the baffles
IHTHs
As Al has said already, many different ways to skin the kitties

I'm not sure I'm comfortable with what some factory engineer have said about rocker geo.
I have it on very good authority that Chrysler believes and says that the stock RB pallet location and curvature completely eliminates all scrubbing motion, and the pallet rolls continuously back and forth on the stem tip.

Except that this is physically impossible.

I agree that 1/3 @ 90 degrees has worked well, but that then just poses the next question: (A) is that how it was designed, or (B) only how it's adjusted, and, if B what was the original pivot position?

Dwayne also has done all my head work in the past on my Kids s/b 360 and my B/B. Had the same issues with geometry and had Mike at B3 make both systems for the 2 race cars. The 360 was the worst on the iron heads with harland sharp rockers I believe we moved the shaft .450 and extended the push rod .500 . I had to add valve cover extensions for the clearances with M T cast covers. The engine has a different sound and has picked up from 11:50s to 11:30s. in similar air. These guys are the best when it comes to service and advise. Thanks Guys.

Mike, I'm curious to know why you didn't make your adaptive eccentric base and cap fit the .750" diameter of the pivot shaft? That would help offset some of the height stack-up from its mounting base. It now fits the O.D. of the adaptive sleeves.

Well, a couple reasons, actually. First, the spacers on the T&D shafts help locate the rockers, not just the hold downs. Second, I pre-machine the materials as much as possible to reduce manufacturing lead time. They are pretty labor intensive, so any changes from the norm is more expensive. And lastly, the roughly.060" shaft spacer thickness being transferred to the aluminum crescent spacer, has no effect on anything stability wise. Six of one, half dozen of another.

Ducati has been using for years, imagine no valve springs and 20k rpm limit. The downside is my 996 Duc is a 90 degree v-twin with 4 cams and 16 rocker arms! Hauls a**, but what a PITA to maintain!