CTSM

This is the first time I do this, so I want to make sure I'm doing this right.

I was testing pushrod length on my LS2 / LSA head combo (milled .03"), I have my factory original LS2 MLS head-gasket which I cleaned up and am using for pushrod length and PV clearance tests.
The final gasket will be a GM LS9 MLS.



I have checking springs on and an adjustable pushrod tester (6.8-7.8), using the actual rockers arms I will be using (Comp Cams 1478-16)



Before trying the adjustable pushrod, I tried the factory pushrod which came with my car (7.397" when new).

I put some dye on the valve stem, then set the intake to the base cam circle (when exhaust started to open).
Finally I set the rocker to zero lash. I cycled the engine 4 times.

Then removed the rocker carefully to inspect the witness mark on the valve stem.
This looks pretty centered to me, what do you guys think? (see below)


Satisfied with what I saw on the valve, I put the intake rocker back on, set it to zero lash again, then I torqued it to 22 ft.lb while counting the number of turns in 1/2 increments.
It took 1- 3/4 turns on the nose. I know 1-3/4 is kind of an upper limit from what I have read here and there, but the mark tells me pushrod length looks perfect.

Since I'm on the upper limit I wanted to make sure the rocker was not opening the valve while on base circle.
So while the rocker was properly torqued, I returned the intake to base circle, this time I shined a strong LED light through the exhaust port.

Held the exhaust valve open by hand (easy to do with checking springs) to make sure that there is no light leaking into the intake port.
I glued my eye to the intake port and no light came through the intake of whatsoever.

So I'm feeling pretty confident about my validation, but again this is my first time don't want to screw anything up.

We get to preload now. The installed lifters, are Katech Johnson 2110-K.

On Katech's page they say:

I'm leaning to Katech's recommendation, question is do I now just add a value within Katech's recommended preload length to the pushrod I used and that's what I will need?

As in if I choose 0.04" as my desired preload (middle point in between 0.035 and 0.045), then the pushrods I need to order should they be : 7.397" + 0.04" ~= ~ 7.44" +/- .01" for measurement error?

RB04Av

Install the heads using the gaskets you will be using; or, if you can tolerate a few .001"s of error, use some pieces of shim stock or the like of the correct thickness to simulate the gasket thickness.

Forget about your stock push rod. Use your adjustable tool.

Solidify a lifter. Molten vaseline or trans gel works great: fill through the oil hole with a syringe like for giving a pet its medicine; allow to cool. Or, if those lifters can be disassembled non-destructively, stack something up in their oil chamber to hold the plunger up, little washers or whatever. Locate peak lift exactly (use a dial indicator; I say "exactly", but it's not super critical, anything within 10° of it is probably entirely "close enough", in that it will make less than .001" to the eventual outcome) on a cam lobe (let's start with, say, #1 int), then turn the crank exactly 1 full revolution, which will rotate the cam exactly ½ revolution and put that lobe exactly at zero lift. Install your solidified lifter, adjustable push rod, and rocker. Exact torque value isn't particularly critical because the whole thing is all metal-to-metal anyway: bolt to trunnion, trunnion to stand, stand to head casting (once it's "tight enough" probably 10 ft-lbs or so on that stuff, you can tighten it until it screams for mercy, but there's nothing to compress so nothing will change dimensionally) but it's easy to get it right so why not. Adjust the push rod out until the VERTICAL play in the push rod and rocker gets taken up. DO NOT attempt to "twist the push rod"!!!! You will screw it up every time. Once that push rod setting is found, observer the push rod and its # of turns, to determine its length. DO NOT attempt to "measure" the push rod with a caliper or any such, you will screw it up every time, because that's not how push rods are specified in the catalogs. Repeat for at least one int and 1 exh at each corner of the motor (1, 2, 7, & 8) since the block decks on the 2 sides, or the 2 heads themselves, might not be the same, and the deck of either might not be "level" front-to-rear; wouldn't hurt to do em all since you're halfway there by that time. Purchase push rods in the length that puts the preload at the max of the range you want to be in (after all, preload almost exclusively goes away as the engine wears: push rods don't get longer when they wear, lifters don't lose play in the roller axles as they wear, rocker trunnions don't get tighter as they wear, cam lobes don't get larger as they wear, etc.). De-solidify the lifter after you're done and before you forget.

You don't need "checking" springs for any of this because your valve train is not adjustable. You're not going to be rotating the motor against spring pressure because you don't have any way of changing anything that you might find out by doing so. Once you get the push rod length right to set the preload where it needs to be, the rest of it just is what it is.

Pattern on the valve stem is irrelevant unless you have some way of adjusting the height of the rocker trunnion. Which I'm not really seeing in your photos. And even if you did, you'd have to set that all up BEFORE selecting push rod length, because any such feature would affect the height of the rocker and therefore the required length of the push rod. Regardless, "centered" is NOT the indicator of best geometry; NARROWEST SWEEP is the ONLY thing that matters. But you have no control whatsoever over that anyway, so there's not even any point in looking at it, beyond raw curiosity.

Push rods come in .025" increments unless you want to spring for custom ones. That's as close as you can get, therefore.

CTSM

I don't have the gaskets on me now, I ordered LS2 gaskets by mistake, they are enroute to being exchanged for LS9 gaskets.
So I'm running the old LS2 gasket. But you raise a very good point, for some reason I forgot about the differences in compressed height in between both gaskets.
The LS9 gasket actually is 0.055" compressed vs 0.051" for the LS2 gasket.

So yes the LS9 gasket supposedly will raise the cylinder head another 0.004". I'm assuming I can factor that in my calculations since I don't have them at the moment?
I was planning on redoing the length check at final assembly anyways.

I have chosen the LS9 gasket over the LS3 because I do intend on supercharging in the very near future otherwise I would have gone LS3 and kept the same compressed height.

That's a valid point too, I just used the old pushrod out of curiosity, since stock pushrods are considered a known quantity to a certain extent and I had them lying around.
For consisetency's sake I have used only one of my old pushrods across all valves on that head and since I had no concerns about bending it, that includes all the valves that do have the BTR springs, the result was the same as the one that did have the checking springs.

Just to make sure I understand you, you are saying that even with checking springs there is travel in the lifter that will skew my results?
I have reviewed every guide I could on this topic, they always say used a solid lifter or checking springs, is there a reason no one talks about solidifying the lifter under test?

I'm assuming I will have to soak it in rubbing alcohol for that to happen?

Additionally, every tutorial I have seen says the adjustable pushrods will most likely bend with high spring pressure (using BTR Dual Platinum springs in this build)

I'm using this method because that's what many guides say to do regarding this measurement.
For example:

RB04Av

Then you can't accurately determine push rod length with certainty. Your result will be however "fuzzy" your "assumptions" about whatever gaskets you're going to end up using, might be.

Good luck with that. The more you "assume" instead of MEASURE, the less likely your end result is to be successful. Note that I DID NOT say "your project will fail", "your result will be wrong", or anything of the kind; what I DID say was that success is LESS LIKELY the more "assumptions" you make. If you're desperate to get this done RIGHT NOW and can't wait to get the right gaskets, then you can adopt an expedient such as shim stock; but then your ultimate accuracy is subject to the accuracy of your "assumption".

Exactly. OTOH, what solid lifter do you know of that's EXACTLY the same height as your hydraulics? I don't know of one either. Note that I DID NOT say "there isn't one"; only, I don't know of one. Maybe somebody that has samples of dozens of mfrs' lifters would know that, but I don't, and I'm betting you don't either. Much better to take one of the lifters you're ACTUALLY USING and turn THAT into a solid, thereby eliminating yet another potential source of MASSIVE "assume" error.

If you use the vaseline method, you don't really have to do anything, although for this purpose, rather than rubbing alcohol (10 - 30% water), lacquer thinner would be FAR better. It'll take care of itself as soon as the motor warms up and melts it. That's the beauty of that particular material for this purpose. You can warm it up and let it drip out before you put it in if you want. Mostly, if you used washers or that sort of thing to do it, you'd need to remember to undo it before final-installing that particular lifter.

I've used an adjustable push rod with the "real" springs on an engine many times - THE SAME adjustable push rod that I've had for around 30 or 35 years - when building motors with adjustable valve trains, where optimizing valve train geometry is possible. However, in these motors, with the stock rocker mounting arrangement as shown in the photo of your parts, that's inapplicable, since the rocker position is not alterable, therefore the geometry is not alterable either. It just is what it is. You bolt the rocker down and you get whatever you get and that's that. Waste of time to even look at it since there's nothing you can do to change it anyway. So since you won't be opening the valves with the adj PR regardless, it's a total non-issue.

You can't. Doesn't matter how "many guides" talk about it. It doesn't apply to your motor. That method applies SPECIFICALLY to engines with an adjustable valve train, as the thumbnail clearly shows (Polylocks), which is NOT what you're building. It is as inapplicable to your project as a bicycle is to a fish, no matter how "many guides" talk about how to ride a bicycle.

CTSM

Fair enough, I was planning to verify my results again once the final gaskets are installed and torqued down.
I realize my build isn't finalized till the new gasket is in and torqued down. But you can say I was practicing, since I'm new at this.

Duly noted sir, I agree.
I was just trying to figure out the ball park figure, I won't be ordering pushrods till I measure my new gaskets are here anyways.

Agreed, there is a lot of assumption baked in here, not arguing there.

Yes definitely would rather not take these lifters apart, they are sold in sets as far as I know, I doubt it will be easy for me to procure a single lifter replacement if I damage or screw one up.
The vaseline method on the other hand, I like that definitely adding it to my book of tricks. It's fully reversible with no mechanical impact, sounds good to me.

That's exactly why I ask questions here instead of thinking I have it right.
I didn't realize that these examples didn't apply to my fixed rockers! Live and learn I guess.

So if I got this right, to put it simply here are the steps I will need to take for an accurate measurement:

- Set an intake or exhaust at the base circle of the cam
- Remove the head, swap out that specific lifter with a solidified lifter (filled with molten vaseline)
- Install the LS9 gasket (will it reach final compressed height with a few bolts or does it need to be torqued to spec with all bolts?)
- Remove the checking springs
- Install the real springs I will be using
- Put back the head
- Install adjustable rod
- Install rocker arm
- Adjust the rod till the rocker tip has zero lash when the bolt is screwed all the way down by hand (not torqued down)
- Repeat for every Int and Ex on every side of the head
- Repeat all above for the second head.
- Note down the measurement or measurements if any differences are detected
- Add 0.045" to the measurement (2110-K max recommended preload)
- Order that length of pushrod either custom or off the shelf (if all measurements are close enough)
- De-solidify the lifter, put it back
- Reinstall the cylinder heads
- Torque down
- Install the new rods
- Bob is my uncle?

BTW, while researching yesterday I came across this video from Melling, I thought was interesting, as they demonstrated a slightly different approach.
The approach they have taken here is to find the zero lash height with the rocker arm torqued, they add half of the lifter travel to the final measurement.
In my case the 2110-k have a 0.200" travel per Johnson, so in my case that would be 0.100" added to the final measurement of each lifter.

Do you think if this process was followed at final assembly for each lifter that the result would vary significantly from your method?

RB04Av

MUCH better. That procedure from Melling applies to LSx motors, unlike the other, which is for SBC, BBC, etc. with adj valve trains.

You use push rod length to set the preload in these motors; nothing else. Unlike those other engines, where you can tip the rocker farther one way or the other at zero lash, to optimize the wipe pattern which should be the narrowest possible regardless of where it lands on the valve stem. These, the wipe pattern is not adjustable.

You don't absolutely "have to" solidify a lifter, if you can be REAL careful not to depress the plunger with your adj PR as you look for zero lash. It's just easier to be certain of an error-free process if you make it impossible to depress the plunger. Note that in the Melling instructions, it says "make sure your adjustable pushrod is adjusted shorter than your expected length": avoiding depressing the plunger is why. Once it's depressed, it might take awhile to restore itself; you never know. You can screw up your measurement (introduce error and uncertainty) that way.

So yeah, you're on the right track now. The Melling instructions are exactly the common procedure used on these motors, and are basically the same thing I would do (or most of the other builder people on here as well). Only thing I would add is, the "half the travel" recommendation they give is really only a "rule of thumb" for stock-ish lifters, such as they sell: others may want some different amount. Even in older SBCs and such, we'd set the preload to ¼ turn or less sometimes, to avoid "pump up", wherein at the upper extreme of RPMs, the lifter plungers would tend to extend too far due to any valve float (bouncing off of the seat) that occurred, which would then cause the valves to get hung a little bit open. The Johnson spec may take this effect into account, or something similar, and want considerably less than "half the travel" accordingly. So use their preload spec, which is .040" or thereabouts if memory serves. IIRC they give an acceptable range, up to that value, whatever it is.

Heads should be fully torqued down so that the gasket reaches its "final" thickness.

With luck the same PR length will work for all 16; or at least, one length for each side. It would suck to have to buy a few each of a half-dozen lengths.

If you're using the Comp adj PR such as the 7702, its threads are ¼"-20, making each turn exactly .050". (1/20 of an inch) With that particular one it's 6.800" all the way screwed together, and you can just count the turns from there to get to PR lengths you can buy off the shelf. Other ones of those may be different but AFAIK all the common ones work on the same principle. No calipers required.

Bear in mind also, the aluminum of the engine castings expands as temperature rises, MUCH more than the steel that the push rods are made out of. This will tend to REMOVE preload (add lash) when the engine is hot, compared to cold. For both block & heads made of aluminum, the effect is pretty huge. The difference in their thermal coefficient of expansion is about 6ppm/°F if memory serves (around 13 ppm for aluminum, around 7 for steel); so over a 200°F range, that's around.009" across a 7.4" span. Meaning, if you preload is any less than that, you run the risk of having LASH when it's hot, instead of PRELOAD. No doubt Johnson takes that effect into account in their recommendation as well. But in any case your preload should never under any circumstances be set to less than .010" or so, cold.

NW-99SS

iTrader: (6)

I would love to delve into this a bit more as the idea of preload measured with a cold engine vs one at operating temp has peaked my interest lately. I know this is over a month old, but thought it might be worth discussing further. Please note, I'm not saying anything is right or wrong, just asking questions to learn and inform myself.

My questions regarding the heat expansion of aluminum vs steel causing a reduction in preload for an engine at operating temps are:

1. In relation to the rotating assembly and valvetrain, are your calculations based on a steel crank, steel lifters and aluminum rods and pistons?
2. I'm also curious about the steel cylinder liners, the use of steel ARP head studs, and how this might affect the calculations.
3. Feel free to let me know where my thought processes/questions are silly or wrong. Again, I'm just trying to educate myself as I go down the rabbit hole of LS valvetrain setups.

Thanks!

RB04Av

Crank, rods, & pistons don't matter to the issue at hand. They might have their own potential behaviors but won't affect valve lash. Pistons for example are well and widely known to cause "slap" until they warm up and expand to their equilibrium running size.

What happens in the valve train is, the castings expand (get larger) at some rate depending on their material, and the moving parts which are steel (mostly the push rods... they're about 80 - 90% of the sum total of it all) expand at their rate. If those 2 rates are different, then the lash/preload changes as the temp changes. Aluminum expands ALOT more than steel, so you gain lash or lose preload with that material. Cast iron expands about the same to slightly less than steel, so usually you lose some lash or gain a bit of preload, but by a far lesser amount than aluminum causes. That's assuming of course that the WHOLE motor is made of the one material or the other; if there's a mix, say an iron block and aluminum heads, then the proportion that they contribute to the overall "height", will affect the end result.

Liners & studs similarly don't matter. Liners aren't involved in moving the valve train at all, and studs aren't materially different from bolts in this respect (both steel) and don't cause the heads to move differently.