Newbie '73T 2.4 MFI Engine Rebuild

[Inkblot]

As previously mentioned, I decided to take a stab at cc'ing my setup, with the goal of confirming (or refuting) my compression ratio math. Again, the issue here is what seems like a discrepancy between other builders' experiences running S pistons in an otherwise stock 2.4T block (the sentiment has been "works great- do it!", and Wayne calls it out in his book as one of his favorite builds) and the basic calcs which suggest doing so would put the compression ratio above 10:1 (which seems to be the consensus threshold for getting into ignition troubles, necessitating twin plugs and/or other anti-knock remedies). In short, are the folks who have done this successfully really running at over 10:1? Or, is the math a little off, and reality is that the compression actually works out to be lower than that?

My goals in cc'ing my setup:

- See how close it works out to the math
- Understand if there are some other dynamics at play
- Do something I've never done before.

I have a set of 2.2S pistons in great condition, which were checked out by my machinist. That said, I know that i could, if necessary, grab a set of lower compression pistons... but I'm hoping I don't have to.

First step was to read through Bruce Anderson's procedure. I have to admit that it took a few reads; the process didn't make intuitive sense to me. Once I understood what was going on, it struck me that there might be a slightly easier way to do it. It's kind of a puzzle; you have to figure out the right combination of using measurements (on the elements which are easily measurable), and doing some liquid displacement tests to either measure the actual volume of something, or the volume around it, which can then be subtracted from an otherwise known/measured volume. There's probably a number of combinations that would work; what follows is what worked for me.

Quick refresher:

CR = v1+v2+v3-v4 / v2+v3-v4

where:
v1= Swept Vol
v2= Deck Height Vol
v3= Cyl Head Vol
v4= Piston Dome Vol

v1 is easy, it's just the cylindrical volume created by the bore and stroke, which are both well-known (and easily verifiable).
v2 took me a while to understand, but it's easy enough to measure.
v3 and v4 require doing the liquid displacement tests, as they're both complex shapes that would be way too hard to try to model out on paper. I suppose if you had perfectly flat-topped pistons (like the stock T's), v4 would be easy, but my S pistons have a serious dome to them.

Measuring the head volume (v3) seemed simple enough: pop in some new valves and a spark plug to close up the holes, and fill it with measured amount of liquid. Bruce Anderson's process calls for making a plexiglass disk with a hole in it, which you can then fill up with antifreeze (for color, I guess), by using burette to carefully measure out the liquid.

When I looked at it, I just couldn't understand why you'd need the plexiglass. It seemed to me that if I carefully leveled the head on a stable table, and used a slightly more viscous liquid, I'd be able measure it out (hopefully without making a big mess!). So that's what I did.

Here I'm shimming up the head to make it as level and stable as possible:




Using a burette (see Gordon's post above for a nice example) seemed like overkill to me. Instead, I bought some cheap plastic syringes from Amazon. With a little practice, I got very repeatable results, and was able to correlate down to the drop.




Also, as you can see, I'm just using motor oil. It worked great. I put a thin layer of grease under the valves to make sure they sealed, and away we go...







You can almost see in this last picture how the surface tension of the oil allows the oil to bulge up ever so slightly above the machined surface of the head. I watched carefully as I filled it up, and stopped as soon as I noticed this effect. Adding any more oil, and it would spill over.




I measured 66.5 mL, which is a little under the reference of 68 I had seen mentioned here previously. This makes some sense, as my heads were resurfaced, which would have the effect of slightly reducing the inside volume. This gave me some confidence that my approach was working...

Next, I'll do the piston domes.


-Jake

[Inkblot]

Getting a volume for the piston dome is a little more complicated. With the crank on my stand, I oiled up 2 main bearing-halves, put them on the crank, and set one half of the crankcase down on top:




While I didn't design the stand to do this, it worked great- I got lucky It's hard to see in the pic, but there was a little gap between the case and my stand (created by the height of the bearings), which I shimmed to make the case as level as possible:




Next, I assembled a piston onto a rod, and then slipped the cylinder over the piston. I lubed everything up pretty good to make sure I wouldn't accidentally scratch anything. I didn't put the rings on, so things slid together pretty easily. Also, the rod (with bearing) is just sitting on the crank- I didn't need to bolt the rod cap on. This setup allowed me to easily turn the crank and position the piston where I needed it (TDC, BDC, etc.).




Next, I set the piston so that the top of the dome was exactly at the top of the cylinder, then applied a little grease along the edge of the piston to seal it up.




Now I just need to fill 'er up.








In this case, the volume of oil applied is the volume around the dome, not the dome itself. Measuring the dome height and knowing the bore gives me the total volume (as if there were no dome), so then I just subtracted the oil volume from that value to get the dome volume. I ended up measuring a dome vol of 25.56 mL. This is a bit lower than the 32mL I'd seen published before, so maybe that's a partial explanation for the "real" compression being lower than the formulaic version.

Next up, final calcs.


- Jake

[David Borden]

I think the issue with not using the plexiglass is not being able to accurately account for the meniscus. You can get a piece of plexi from almost any hardware store, and my guess is its done that way by the experts for a reason.

I like your willingness to try something outside of the box though. It would be interesting to see you attempt the tried and true method and see if your results match.

[Inkblot]

When thinking through all of the compression ratio stuff, I found it helpful to build a visual model, so that I could better grasp what the formula was actually doing.

Here's a quick mock up:




...and the individual pieces:




Measuring the dome height is critical for calculating both the deck height and the dome volume. Aligning calipers on the shoulder of where the dome meets the cylindrical body was unreliable when I tried to measure the dome height directly. I had an easier time just measuring the total height and subtracting the cylinder body length to get the dome height:




The dome volume is figured via subtraction as well. The purple shape is the volume as if there was no dome, which is easy enough to calculate using just bore and dome height. The brown shape is the measured volume of oil that fills in around the dome; subtract that from the purple value and you get the dome volume. Easy 'nuff!






- Jake

[Inkblot]

So, just to wrap up the CR formula graphically:




Doing all of the measurements was very informative. Here's what I ended up with:



The table on the left has my raw measurements. I got 10.11:1, which isn't far off of the 10.39 the formula would suggest. At closer look, there are some interesting differences:

- My piston dome vol is noticeably less than the values I'd seen previously published. It's important to point out that I can't verify the published numbers as being accurate.

- My head vol is only slightly less than the published number, which makes some sense as my heads were resurfaced, thus removing a small amount of material, making the internal volume a bit smaller.

- My stroke measurement seems right on.

- My deck height seems really low. This also makes some sense, given the machining on the heads and block, both of which would have the effect of moving the head ever so slightly closer to the block. I measured this using a single, 0.25mm thick (need to confirm this) copper gasket (aka shim?) between the block and the cylinder.

The good news is that by installing a thicker gasket, I will both increase the deck height to a more comfortable value, while reducing the compression ratio. Wayne's book suggests that you shouldn't run a deck height of less than 1mm, so that'll be my target. I gather that you can find shims in a number of different precision thicknesses, so I'll just have to see what's available, and err on the safe side.

Looking at the table, if I bump the deck height to about 1mm, my resulting CR would be 9.4:1, which sounds great to me.

I'll track down the new shims, and then I'll re-cc and see where I land.


- Jake

[Neil Harvey]

Use a thinner fluid. That will make a difference. Oil is too thick and will skew your volumes. Try a cleaning solvent with a little Red ATF added.

[Inkblot]

I gave it a second go with some coolant, just to try a slightly thinner fluid. I think the real issue here is surface tension, as opposed to viscosity- so a solvent or something like that would probably yield even better results.

With a few practice runs, it's actually pretty easy to gauge visually how much of a bulge you can create, and to back off until it looks just right. Here are some examples:













I found that the surface tension of the coolant would allow me to add an "extra" 1-2mL. After getting everything to look how I wanted, I ran the calcs again. I wasn't significantly off from my first attempt with regular motor oil, but my low deck height value was still bothering me. As I mentioned before, it makes some sense that my value would be lower than stock, given the machine work- but I wouldn't expect that much material to be removed.

I decided to remeasure and redo the deck height calcs using a more appropriate measuring tool, so I picked up a cheap height gauge:



In this pic you can see that the positioning of the gauge is a bit futzy, so I did it again with the gauge up on blocks (of known height) to confirm my value. I also tried turning the gauge 90 degrees, to align it with the piston pin as some have suggested; didn't make any difference.

I got a noticeably different measurement this time around, from 0.343 to 0.495. Might not seem like much, but this makes a significant difference in the resulting compression ratio calculation. After scratching my head for a while, it all started to make sense. The measured height in mm of any of the volumes here is a much more significant value than any of the measured volumes themselves, by roughly 5:1. At a bore of 84mm, 1mm of height results in 5.54 ml of volume.

In other words, it's much more important to get your height measurements as precise as possible, and the volume measurements with the coolant or whatever don't have to be as perfect. I was approaching it the other way around; I was trying to get my syringe measurements down to 0.1mL precision, when I should have been focusing on my height measurements instead.

Another way to illustrate this is to relate the measurements to the resulting compression ratio. For example, holding everything else equal at 84mm bore, every 1mm height changes the CR by 0.89, whereas every mL only changes the CR by 0.168.

So, I think any concerns about measuring the meniscus or surface tension or whatever are largely overblown; you can be off by an mL or more without a huge impact to the calc; but if you're off on your height measurements you'll see a big change in the resulting CR.

I'll share my updated calcs next...

-Jake

[Inkblot]

I picked up some new shims from the helpful folks at EBS. I got a set of 0.5 and 1.0mm, to add to the 0.25mm "stock" ones I already bought.



Running the calcs for a target of 1mm deck height would require roughly 0.75mm shims... While I could get 0.75 by stacking my 0.25 and 0.5's, I've read that it's less than ideal to stack the shims- so I'll go slightly conservative and use the 1.0mm shims. This should yield a final CR of about 9.2:1, with a safe deck height of 1.2mm. Sounds good to me!



The first column shows what I'd get with the 0.25mm shims, the second with 1mm.

I'm now feeling confident about using my 2.2s pistons, so I can get back to balancing and assembling the rods.

- Jake

[Mixed76]

Well done testing your method and rechecking results. Great post.

[Neil Harvey]

Question, when you had the piston at TDC then lowered it ( measured amount) to find the dome volume, did you add in the piston deck height into the amount lowered? Did you add into your CR calc the volume above the top ring to the piston deck. This the diameter above the top ring and the distance from the piston deck to the top pf the top ring.

Doesn't make huge difference, but it all adds up. Especially if you are trying to run as closed as possible CR to detonation limit.

[Inkblot]

@Mixed: Thanks!

@Neil:

Originally Posted by Neil Harvey

Question, when you had the piston at TDC then lowered it ( measured amount) to find the dome volume, did you add in the piston deck height into the amount lowered?

Not exactly... I didn't lower the piston by a measured mount, per se; I lowered it until the top of the dome was flush with the top of the cylinder. This way, the dome height, which I had measured separately, defines the volume around the dome.

Originally Posted by Neil Harvey

Did you add into your CR calc the volume above the top ring to the piston deck. This the diameter above the top ring and the distance from the piston deck to the top pf the top ring.

Doesn't make huge difference, but it all adds up. Especially if you are trying to run as closed as possible CR to detonation limit.

Nope, didn't do this. I greased up the piston, but didn't put the rings on. The grease sealed around the piston to keep the fluid from leaking out, but you're right- it created a barrier at the piston-deck edge, not the ring edge. I think it's probably a very small amount, and given that that volume is there at both TDC and BDC, it probably has only a minimal effect on the final CR- but you're absolutely right that it's not in my calc. I'll see if there's an easy way to quantify it...


- Jake

[Inkblot]

@ Neil, thinking through your question a bit more: the small amount of volume created by the distance between the piston-deck edge and the top of the ring would have the resulting effect of increasing the head volume (V3 in the CR calc) by some amount. Increasing this value decreases the resulting CR, so by not counting it I'm "playing it safe" with regards to the CR. I can't claim to have done this intentionally or intelligently, but that's the result
So, I think you're correct- especially if you're trying to maximize your possible CR, it's probably worth taking into account. At my 84mm bore, every 1mL of volume is equal to 0.18mm of height; so I suppose you could then recover that CR loss by decreasing your deck height (for me, by using slightly thinner shims) by the appropriate amount.
In my case, I was worried about the CR being too high- so knowing that my real CR may be slightly lower than my calcs is a-ok w me.

Also, the volume in question should be easy to calculate, if we know the height from the top of the ring to the deck edge, and the gap (measured or spec?) between the piston and cylinder wall. But I don't have those handy at the moment....

- Jake

[Neil Harvey]

Jake,

Well done. Glad you understand what I was talking about.

Often when measuring the dome volume, the piston is lowered a certain amount and this amount is used in the total volume calcs forgetting about adding in the piston deck distance.

As for the volume around the top of the fire band, we have to use this especially when running right up against the rule book or detonation limits. This get larger the bigger the bore.

[Inkblot]

@Neil- yeah, makes sense that this would have a greater effect at higher displacements.

Back at the shop, I ran a few numbers:

ring to edge h (in) 0.1968
piston d (in) 3.2164
cyl d (in) 3.304
gap vol (in^3) 0.088286301
gap vol (mL) 1.44675644

So, it looks like the gap space between the piston and the cylinder wall, from the top of the top ring to the "corner" of the piston (the edge where the dome starts) creates a volume of about 1.44mL.

From before, each 1mL is equivalent to 0.18 of height or 0.168 of compression ratio. Therefore, the newly measured gap volume should subtract 1.44 * 0.168 = 0.243 from my previously measured CR, giving me right about 8.95:1 (down from 9.2:1).

Correspondingly, I could decrease my deck height by 1.44 * 0.18 = 0.26mm to compensate. This is interesting, because I think I have the right shims to accomplish this...

Here are the options, as I see them:

A. 1mm shim = 1.2mm deck height = 8.95:1 CR

B. 0.5+0.25mm shim = 0.95mm deck height = 9.185:1 CR

C. 0.5mm shim = 0.7mm deck height = 9.43:1 CR


Option A is the safest, and though it may leave a little performance on the table, it should be still noticeably better than stock.

Option B looks like the sweet spot between CR and deck height, except that it requires stacking 2 shims. I've read mixed opinions about whether or not that's a good idea.

Option C looks doable, and is closer to what my original target CR was, but that deck height is a little on the low side. Might still be fine, but would require that I'm very careful with testing my valve-piston clearances.


Any opinions out there? I'm leaning towards option A, unless anyone has strong opinions about the viability of a 0.7mm deck height.

Thanks all-

Jake

[cmcfaul]

From the start of this thread I was going to recommend the .5 mm shim. That's about the ideal CR and gives you .5 margin of error.

To be safe, use a slightly colder plug. Don't run too lean. Always use the highest octane gas. Might want to add an external cooler like the 911S and later models to keep temps down. A
more aggressive cam like the mod S or DC 30 is ideal as it has more overlap which takes advantage of the higher CR. From experience, an AMAZING combination. standard cam runs out of breath at 6k RPM. Mod S or DC 30 hits the power band at 4 k and will fly past 7300 RPM redline pulling ever harder the higher you let it go. Its the engine Porsche should have built.

Chris
73 911 E

Chris

[Inkblot]

Thanks, @chris. I'm still trying to understand the relationship between the deck height and the piston-valve clearances, and I gather that the only way to know for sure is to assemble everything up through the valve train and then measure and test. I like the idea of putting modeling clay on the piston top, which gets deformed by the valves in a way that can be directly measured; but I haven't decided yet. I'll be running E cams, and I have the MFI pump calibrated for them. My understanding was that the more aggressive cams wouldn't do a lot for me unless I opened up the intake ports a bit.

Jake

[Trackrash]

Yea, to be sure, you will have to assemble and measure. However, I don't believe you will have an issue with piston to valve clearance using E cams and S pistons. But none the less, you need to check.

If you think about the way the motor works, the piston will chase the exhaust valve during the exhaust stroke.

As far as port size goes, I agree, opening them up will help the top end even with E cams. If it was me, I would go up to 34mm on the intake. No reason, IMO, to go further even though the S had 36mm ports.

If considering more cam, a mod solex grind would be my choice. If you go with 9.5 CR you would have a well balanced motor with a good torque curve.

That said the 2,4 E had 32mm ports originally.

The issue is your throttle bodies will need to match your ports.

Don't discount a sport muffler and K&N air filters. Those can add 10 HP easily.

[Neil Harvey]

I have suggested on this forum how we go about checking Piston to Valve clearances. We don't use clay nor do we use the rocker arm method. We do it how we do, so we can on paper establish a piston deck height, any cam we may choose and any setting that cam may use.

Again,

Dry assemble the engine with one piston and its cylinder with whatever base gasket you may choose to use. If you used a certain thick gasket to do the CR check, use the same one. Fit a degree wheel to the crank and TDC the piston. You should split the dwell either side of TDC to find true TDC. Turn the engine to 20 BTDC. Fit a Cylinder heads with the valves loose in the guides. No springs. They will drop down and contact the pistons. You can hold them up with rubber bands if you wish. Bolt the head down. Its not necessary to torque the head nuts. Once they are snug the heads is down. Do the following 1 valve at a time. Fit you dial indicator and the pointer in line with the valve stem and touching the top of the valve. Allow enough travel from the seat to the piston.

Pull the what ever valve up against its seat, zero out the dial and slowly lower the vale until it touches the piston. Record this distance. Do the same every 5° until you have reached 20° ATDC. Now do the same for the other valve. This now gives you the maximum distance from the valve on its seat to the piston. Its always good idea to make sure the valve go into and has clearance to any pocket as well.

Now you can choose a camshaft, any setting and calculate the clearance. You can also figure if you need to add or remove base gaskets. Any good engine shop or even you cam company of choice can tell you or measure your cams and tell you the actual valve lifts at each of the pistons positions you measured, at whatever setting you may want. For example, you want to use, a particular cam but are unsure if it will fit. Have the cam supplier tell you the suggested timing positions and ask them to give you the valve lifts at those piston positions. Now all you have to do is simple math to find out your clearances. As long as you use the same base gaskets you numbers will never change. If you change the base gaskets, its again a simple math calc, trig actually to calculate the differences, or just remeasure.

This may seem long winded, but its very easy and makes cam choice easy and if you have an issue its all on paper. And yo don't have to tear your engine back apart again.

If you have any questions, email me.

[Mixed76]

Nice and simple Neil, and sounds bulletproof. Thanks for reposting.

[Inkblot]

@Neil, thanks for the well-written instructions. I'm not there yet, but they will definitely come in handy!