Elementary Q&A - First Time Engine Builder 4

[MichaelPGriffith]

I would like to open this thread and keep it as the place that I come to seek advise on an engine build I am in the process of, without cluttering the forums with repeated elementary questions.

I'm working on a naturally aspirated build; Honda J series V6. I'm doing everything that can be managed outside of a machine shop by myself; including thorough massage of the OEM intake manifold and heads; long tube stepped headers; and many other things that a completely inexperienced person such as myself probably has no business attempting.

Here is a consolidated album that shows the current progress of my project:

http://picasaweb.google.com/SoEasyAuctions/MR2Consolidated#

At this point the intake manifold is coming along quite well and I'm getting ready to start on the heads; which leads me to my questions:

1. Where can I find some good layman's reference material on port shape and theory to help me decide how to attack the ports? (If this doesn't exist, maybe just some pointers?)


2. What would be the best method of determining the clearance between the piston @ tdc and the chamber? I want to increase the quench area.

 

[MikeHalloran]

1. Just remove the imperfections, if any, from the factory ports.

2. Modeling clay.



Mike Halloran
Pembroke Pines, FL, USA

 

[MichaelPGriffith]

Thanks for the reply.

These engines are producing ~300hp @ the crank from the factory, so I think the head is optimized for ~395cfm.

A few guys run them with basic bolt-ons and are seeing ~300hp @ the wheel; ~340hp @ crank; ~430cfm.

Both of these examples are making this power at ~6700rpm.

I'm trying to make peak power @ 8k, and if I plug the same VE #'s into a simulator I'm getting a CFM @ 8,000rpm of 490.

Again, I'm a dilettante when it comes to these things, but that CFM is almost 25% over original; I would think that some amount of port shaping will be necessary to avoid having a bottle-neck there.

No?

 

[MikeHalloran]

How many heads do you suppose the factory trashed during development?

How many can you afford to trash?



Mike Halloran
Pembroke Pines, FL, USA

 

[MichaelPGriffith]

Well I actually have two sets, but I'm really hoping I don't have to trash any. I get your point though.

My thoughts were to leave to floor depth the same, add 1mm around the sides, widen the throat a couple mm's @ the split point, blend the valve seats, and then lightly radius the bowls.

I've read though some things on the subject and am quite aware that there are certain areas of the port that shouldn't even be touched. I was just wondering if there were some simply theories I could put into practice, but it seems you're telling me that actually doing something beneficial would take much trial and error.

 

[patprimmer]

On a factory modern high performance engine the factory has already done the simple theory stuff and some complex theory stuff with computer simulation.

What the factory may or may not have done is made some inperfections due to machining and casting economic limitations. These may be casting flash and sharp edges and details in the area of the valve seat and valve guide and surface finish.

On a high performance Honda engine I would simply remove obvious obstructions and blend the sharp edges. I would open the chambers where the valves are shrouded by the bore, but I would only take that to match the bore.

There may or may not be a bit to be gained by shortening or tapering the guides.

A satin finish on the inlets, especially the roof or outside rad helps a bit. Keeping the inside rad concentric with the outside helps a bit but most flow happens on the outside.

Polish on the exhaust and especially the chamber helps with reduced heat transfer from the charge to the head. This is especially important in the chamber where all sharp edges that can form hot spots that induce preignition or detonation should be smoothed out. Same goes for piston tops.

It is also important to have good tight quench clearance.

This is controlled by a trial assembly where you measure how high the piston top is compared to the block and how thick the gasket is and if the quench area on the head is at the gasket surface or recessed into the head. If at the ground surface measurements surfice, but if recessed, you needclay. You still need clay to measure piston to valve and the closest points ARE NOT TDC. The exhaust is closest BTDC and the inlet ATDC. Vtec also has a strong influence on piston to valve, so you need to check with vtec engaged.

If you intend playing with cam advance and retard, you need to take the extremes into account when testing piston to valve.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[Warpspeed]

Yup,

These days high performance factory heads already have the correct shape and dimensions all worked out.
All they lack is the final hand finishing and cleaning up of any really obvious casting imperfections.

If you think you can do some really radical things, and vastly improve on what the Honda engineers developed after a huge amount of very expensive long term testing and development.... think again.

Just clean up the existing ports, make sure the valves, valve seats, guides, and springs are as good as they possibly can be, and concentrate on other more productive things.

 

[MichaelPGriffith]

Thank you gentlemen for bringing me back down to earth.

Any advice on determining a good length for the headers I am going to start on in a few weeks? The factory manifolds have ~12" long primaries, and that seems to be the pattern even with headers.

I was going to try a best-guess length of ~36" long primaries.

 

[MichaelPGriffith]

Also, Pat, do you recommend polishing the dome, chamber, & exhaust port over ceramic coating?

 

[Tmoose]

On a more conventional engined I'd think retarding the stock cams would not be enough to move the power peak upward 1300 rpm.

By "increase the quench area" do you mean increase the clearance between piston and cylinder head? For effective quench/squish for detonation resistance it needs to be ~ less than 1 mm or so.

 

[evelrod]

So far so good. Just clean up the ports and port match the manifold. I agree that there is a little to gain by re working the valve guides, but not a lot...AND...that is a point where you can make it radically WORSE! I disagree with Pat on the port finish...we have found through many examples that a highly polished finish is detrimental to flow and carbon buildup...just a nice smooth satin finish. Also, the several methods of coating the combustion chamber and piston top have proven their worth in ultimate builds. $$$$$$, though! The best bang for the buck is a good blended (multi angle) valve job and careful matching of springs/retainers, etc. I'm a bit anal about "balancing" my valves/springs/rockers/pushrods/tappets...leaving "no stone unturned" approach.

For quench/squish I try for as close a tolerance piston/cyl.head as possible. My Mini is ~0.025" max and the Lotus is in that ballpark, too. That may be a bit close for a street engine that is not periodically "de coked". Piston to wall clearance will be specified by the mfgr and should be adhered too. Using a torque plate for final honing of the block is a must as is honing and, at least, checking all other block and head dimensions (you would be surprised at some of the "factory" mistakes). Using top quality pistons/rings with a zero gap second is good in many cases. I use one millimeter coated top rings.

Piston to wall on the newish hypereutectic alloy pistons can be held to extremely close tolerances thus making ring sealing much better. Care should be used with these pistons if you are shooting for an all out performance engine. A good forged piston will require a bit more clearance but, is a bit stronger for a supercharged, etc. engine. Some of the imported car magazines claim over 1000 hp on stock bottom ends...Hmmmmm?

You cannot get a crank journal polished enough to suit me...;o). I still use more clearance than most @ ~0.0015" to 0.0025" rods and mains with a good 40 or 50 Redline racing oil.

There are so many publications out there I don't know where to start...In 55 years, I think I've read most of them. All good in one way or another. You just cannot get enough information...ever!

Rod

 

[MichaelPGriffith]

Tmoose; when referring to increase in quench area I am referring to a decrease in piston to chamber clearance. I understand that the last moment "squish" of charge that had occupied the open space BTDC is beneficially agitative, and the resulting denser a/f helps provide a faster combustion event.

There are no cam options out there that could fall within my budget; shelf units just don't exist. To help the engine take advantage of the larger intake runners (25% cross sectional increase @ entry & ~10% increase at port entrace), and headers ect., I am bumping the compression ratio from 11.2:1 up to 12.8:1, perhaps a little bit higher.

The factory ECU's start pulling ignition timing @6,800. Do you think the bump in compression, better manifolds, and standalone EMS are capable of achieving peak power @ 8k?

There just aren't many naturally aspirated J series builds out the. In fact I don't think any exist that have gone any further than basic bolt-on parts. There are a few nice FI builds out there, but they don't tell me anything about my goals.

ROD: Thanks for the overview. I know that I am going to go with stiffer valve springs and ti retainers. The pistons will remain the OEM cast units, but I have smoothed the dome and removed all casting markings and irregularities. I will make some slight modifications to them; bridge the valve pockets, ceramic coat the domes, dry film lubricate the skirts, thermal dispersant coat the underside, and reem and radius the oil channels; they also need to be notched to clear the oil squirters that are being retrofitted.

Would you recommend that I stay away from using new OEM rings in favour of something aftermarket?

 

[evelrod]

For the OEM pistons, OEM rings would probably be fine. I cannot make a blanket recommendation as, in my experience, each piston mfgr prefers a different ring...That, and, of course, each engine builder has their own preference. This is what's in the Mini right now, but in the past I have used Deves with good results...Lotta guys disagree with me on that, though.

With stock cast pistons, take care with the compression ratio. I use 15:1 on the Mini, but not with cast.

In my experience, admittedly a long while back, a good "blueprinting" (old terminology) would yield perhaps 20%. I'm not sure you can push for an additional 2000 rpm. I'm betting that's just wishful thinking, given your design parameters.

Rod

 

[patprimmer]

Polishing is cheaper if you do it yourself. Ceramic coating is more durable and effective.

I would still do a very smooth satin finish before coating as this reduces surface area and therefore heat transfer out of the chamber. As they sandblast before coating to get adhesion, better than a satin finish is a waste of time.

Looking at the photos in detail, I would not touch those guides. They are already pretty good and may be acting as flow directors to direct flow toward unshrouded areas.

The head has recessed quench pads. A cut on the heads could improve those pads. How thick is a stock head gasket?

I would aim for about 0.030" deck clearance. The bigger the bore, the higher the rpm, the more flex in the crank, the bigger the piston to bore clearance, the more deck clearance you need to accommodate rod stretch, piston expansion, and the piston rocking in the bore.

There might be something to be gained in the throttle body/plenum region as these are quite likely compromised for packaging reasons as this engine was basically designed to fit a V6 for the American market where a I4 normally sits for the rest of the world.

Also, Tmoose is correct. Retarding the stock cam will not be enough. To get the most you need a cam with more duration and lift.



Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[MichaelPGriffith]

Again, everyone, thank you for the input.

Pat, the factory HG is .75mm thick and three layers. I intended to only run a single layer of it, but it sounds like I can make sure the dome sits flush with the top of the bore, fill the quench pad with weld, and then have it shaved a few thousandths.

The internal coatings were also going to be a home project; Techline coatings. Familiar with these? Good quality?

The intake manifold is several hundred cc's larger than stock since I have removed all of the EGR valving and opened up the runners. I haven't gotten to it yet, but the throttle body is going to be a 90mm LS1 unit. The neck that the throttle body mounts to doesn't even have a 90mm OD, so I will cut it off and weld a very thick adapter plate on; something tall enough to give a nice gradual radius to the inlet.

I've known all along that the perfect intake and exhaust track wouldn't be worth a hoot if there wasn't a need for their increased flow ability. I was really hanging my hat on the bump-the-compression-up-significantly idea. I understand that to get "the most" out of the whole system I need cams designed for my intended power band, but, without custom cams, do you think the engine will just fall off at 6,800rpm? Or do you think that the other factors being changed will contribute to it making more hp @ 500-1,200rpm above previous redline; granted not linear and as significant as it could be?

That said, best guess, best case scenario, what kind of cost should I expect to have a bespoke cam turned? It seems I might need to hang my hat on that.

 

[Lou Scannon]

Assuming changing the cams isn't extremely difficult after the fact, you might consider building with the stock cams, and holding back the cam upgrade for the future.
I think everyone here has known the irrestistible urge to tinker following a build that was supposedly "completed".
It would be a good experience for you to assess the improvement from the porting & compression ratio mods (which are difficult to do after the fact), and subsequently assess the delta improvement from a cam upgrade.

 

[patprimmer]

Techline coatings are fine.

Do not weld the head. Mill the head and the deck of the block if necessary and run the 3 layers in the gasket.

You may need adjustable cam gears to correct for the cam being effectively moved closer to the crank and to move the peak power point up the rev range.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[Tmoose]

Is this a turbo?

Dan T

 

[MichaelPGriffith]

No it is not a FI project as of yet. I intend to eventually build a bullet-proof bottom end for it and run a pair of large Rotrex superchargers, but that won't happen until I have the chassis in a state capable of making use of that much power, and the rest of the engine, fuel, coolant, and electronics systems up to snuff.

 

[Marquis]

It took me many years to develop a series of port shape modifications for the BMW M20 series of cylinder heads. I had access to a flow bench

see here

http://speedtalk.com/forum/viewtopic.php?p=83856&sid=3fe13da6c51b11381c3c2ac64713b4b8

And I went through several cylinder heads for development. The advantage I had is that these are old engines quite often with cracked and porous cylinder heads so I would pick up cracked castings for about 20 quid and section them. I would then make various mods and test them. In addition I had 'male mouldings' of the ports made using silicon rubber and visualise what I was doing. I ended up looking at the valve guide length and evaluating it as a percentage of the total valve length- it was about 51-53% while modern engines are getting closer to 40%- so I reduced this and raised the port roof area. I also ended up spending alot of effort on the short side radius. I came up with an exceptional port. For repeatability I formed some 2 D templates- not disimilar to these:

http://www.moparmusclemagazine.com/techarticles/5115_cylinder_heads/index.html

I'm now undertaking this whole process again for the old Big Block Dodge engine. However the Dodge engine really is leagues behind anything made by Jaguar, BMW, Alfas that I'm used to so I have my work cut out for me. I have access to a flow bench and scrap cylinder heads. Without access to these I would make 'male mouldings' of the ports- several infact and section them to take planes and get sections through the ports. I would then visually inspect each of these sections through the male mouldings of the ports and make an engineering judgement of what needs to be done to smooth out the flow from experience. Without a flow bench you're really flying blind. Also- a modern port such as a Honda- you'll probably find there's very little that needs to be done (unlike a 1960s Dodge). You really STILL do need at least one scrap cylinder head to get an assessment of the thickness of the metal sections and where the cooling passages are.

http://www.auto-scape.com