Nicks Diy Head

[Nick Jones]

This is a follow-on to

http://sideways-technologies.co.uk/forums/index.php/topic/690-diy-heads/

and should be read in conjunction with it.

 

Didn't want to add this onto the bottom of Neils excellent thread, but make no mistake, I'm drawing heavily on Neils information. I've added some observations of my own too.

 

The victim is the head from my Mk2 2.5PI. FWIW the number is 218225 though for the purposes of this modification they are are pretty much the same apart from the 219016 head from the 2.5S which are quite different in the inlet valve throat area.

 

What will vary from head to head is the amount of core shift. What the hell this that? Well this head is quite bad, so I'll show you.......

 

When the heads are cast, after the main impression is made in the sand which form the basic shape, cores are laid in that will create the voids for the waterways, ports and the basic shape of the combustion chamber.

 

The port cores want to be positioned so that the centre of the section that forms the valve throat is concentric with the position the valve guide. That means that when the valve throat and seat are machined later you get a reasonable blend of shapes without any serious steps.

 

The pic below shows that the cores on this head were too far to the left so there's a pronounced step in the short side radius and an undercut below the seat area on the far side.

 

 

This one shows the undercut better. This is the bigger problem really as the SSR step can be cleaned up, but this really needs material added.

 

 

About all you can do with this is to chamfer the sharp edge. You are limited by the need to leave plenty of material under the valve seat. DO NOT undercut the seat!

This is my attempt

 

I don't know how big an effect on flow this actually has.

 

Another interesting thing noticed was the evidence for high and low flow areas in the valve throat.

 

 

Yellow arrows show the cleaner areas indicating higher flow,

Red arrows show heavier deposits indicating low flow. The section just next to the exhaust valve appears to have the lowest flow by some margin.

Green arrow shows what appears to be the "flow shadow" of the valve stem. Never seen this before and I think it's partly due to this being an injection engine. It's not equal on all cylinders, which may be telling me something about the state of my injectors.

Blue arrows show areas of heavy deposits that always seem to form in these corners indicating stagnant areas.

As a general comment, this head came from an engine which was a VERY heavy oil user. The flow patterns don't show through into the combustion chamber all that well. This may be because I wasn't using the engine hard before it was dismantled.

[Nick Jones]

So, to the port work.

 

I've done these before. Last one I did was the Vitesse one, with a biggish Dremel-type thing (I used it here for the finishing work and it's shown below).

 

This was £ 30 from B & Q several years ago complete with the flexible extension and a good selection of accessories. It's a good tool and very versatile. I've bought a few small carbide burrs over the years but it still takes AGES to remove much material. So this time I treated myself to a proper electric die grinder which is something I've wanted for a long time and will use for many things. Cost £ 85 delivered, which I thought was pretty good value.

 

 

I also bought some stones of various shapes from these people.

http://www.abtec4abrasives.com/mounted-points-wheels-and-burrs-for-portable-machines-57-c.asp

Not used them before but I will use them again

Of the stones pictured, I only actually used the round one - this one

http://www.abtec4abrasives.com/a-25-mounted-point-ra-60-rv-685-p.asp.

 

It's a little bit big - not a problem for me as I wanted the guides out anyway but if you want to leave the guides in, choose the next size down.

 

I also used the carbide burr that's shown in the grinder. I've had that in a drawer for years. Note that it is a "bullnose" one and not really the right shape for this job as it cuts on the end as well, making it all to easy to dig ridges as you are cutting right around the curve. The one Neil shows is a much better shape for the job.

 

Also in the first pic you'll see the visor, earplugs and dust filter. What is not shown is the vacuum cleaner which happens to have an attachment which wedges nicely into the ports and was left running while I was grinding.

 

After the small tool, the die grinder and carbide was quite a revelation - take it careful or you'll be through to the water before you know it! I was also paranoid about taking a divot out of the valve seats so made a policy of not starting the grinder until the tool was in the port and not taking it out until it had stopped turning.

 

First cut as per Neils guidance and measured with homemade depth gauge similar to his. You also get a sense of the size and shape of the core shift step in this pic.....

 

With the carbide I was using I had to be careful not to dig too deep at the back and make a step. Having set the datum I then put a permanent marker line down the deepest point of the cut to try and stop myself digging any more out of there and used the carbide to open out either side, with the deeper, more laid back angle biased towards the outside of the chamber. The carbide is a bit fierce so I soon swapped to the round stone as it's wider contact patch and shape make it easier to form the gradual, smooth curves that a needed. Just kept it moving back and forth around the shape of the curve, using only light pressure and pretty quickly got to this

 

Note the deeper cut with the shallower angle biased towards the chamber wall. Hardest part was getting a smooth blend up round the bend - partly due to the carbide having left ridges initially I think.

 

Then polished up using the sanding drum on the small flexitool and got to this

 

Quite a marked contrast to the unmolested one next to it!

[Nick Jones]

On the exhaust side I restricted myself to chopping the bulge in the guide boss away and smoothing the short radius which was also very sharp due to the core shift. I haven't moved much material as the head will have inserts fitted for unleaded/lpg and I was warned these will be quite deep.

 

 

Likewise, on the chamber itself I've just radiused the corners where the sloping sections of the squish ramps meet the vertical chamber walls.

 

 

I'm not expecting this to have much effect on flows but I'm hoping it might aid a more complete burn. Not much material removed - less than 5mins per chamber with the small grinder and fine carbide burr. I did this much more aggressively on the Vitesse head and had trouble getting the CR back up. Not such a problem on this head as much less material removed and 2.5 so more spare material. In any case I don't mind a smaller chamber on a 2.5 as this adds squish area and I think this helps the engine run at higher AFRs. I've not yet had the head off the Vitesse to see if this treatment has reduced the deposit formation in the corners.

 

Next job was to fit the new guides. I just bashed the old ones out with a brass drift and lump hammer but fitting needs more subtlety. I've got a small press so I made up a little tool out of an old valve stem and piece of 8mm strip plus cut a section of box to 16mm as a stop to set the insertion depth.

 

 

This worked fine on the inlets. However, when I got to the exhausts I discovered the 5 out of 6 are undersize and can be fitted by thumb pressure or even gravity.......

 

More crappy parts - see post in rant section for vid of valve guide going in under gravity!

 

Cheers

 

Nick

[Triumph-V8]

For my convenience I added a circumfencing ring area around the guide with a lathe.

This area accepts a round ring (Sprengring) and is positioned that it is flat

on the ground of the head when the guide is in position.

First you get always the proper depth

second the guide can not drop when head is getting hot.

The ring remains in that position and does not interfear with the spring(s).

 

If one wants to make it perfect

one might want to drill a little 45 degree angle

at the head in hole for the guide.

[AlexC]

Looks great! I miss doing this stuff. I was so obsessed when I did mine. Might be worth removing the spark plugs if your going to reuse them though. After pressing in my bronze valve guides I had them reamed to spec as they because very tight on the valve, even with a little lube. I remember someone saying that wouldn't seize with the bronze and would wear in with a tight seal had I left them.

[Sam]

To finish off the inner radius, I cut a long thin strip of sandpaper off of a beltsander, and used that to do the final shaping - you can really tug on it to flatten out the lumps.... it can be hard to get it smooth with tools poking in from each side.

 

Your finished product looks very similar to mine - I think I removed a bit more material on the shoulders, and mine is a rougher finish on the inlets...but its a1500 head so a bit different anyways..

 

Is polishing the exhaust ports a good idea? It's not like the inlets where fuel has to atomise or anything.

[Wolfrace.]

 

 

Is polishing the exhaust ports a good idea? It's not like the inlets where fuel has to atomise or anything.

 

Just clean it up and match it to the manifold, within miutes of starting the engine it will be sooted up no matter how mirrorlike the finish.

[Nick Jones]

Guides obtained from Peter Burgess, inlets modified so I can fit seals if I want to (with double springs). These were plenty big, allowing a little light fettling to get the fit of your choice. Will also need a hone to get the right fit for the valves. This will be done by the guy who does the seats.

 

Now have most of the bits to build the bottom end. County pistons were all within 0.5g for weight except one which was 1.4g heavier than the lightest. Bloody baby grinder blew up (literally - flash! Bang!) while relieving the inside of the gudgeon pin to correct this so I'm stuck again......

 

Con rods have a 5g spread. I'll be attending to this.

 

Nick

[Wolfrace.]

Alas, metallic dust and electrikery are poor bedfellows, and cast iron dust is particularly bad with electrics.

[Nick Jones]

You are not wrong...... especially when the vents in the casing direct the flow in such a way that it deposits a stripe of metallic dust right across the narrowest point between two exposed tracks on the pcb..... arc - kaboom..... eeek! bollocks!

 

Cleaned it off, mended the vaporised track, replaced fuse......... still dead. Something else is kippered as well. Not much else on pcb so might still manage to fix it..... I have another but it died the same way some time back and has already donated some mechanical parts to this one.

 

Wasn't even very near the work area this time as I was using the flexidrive and had the hoover going too....... B & Q seem to have stopped selling them.

 

Nick

[Wolfrace.]

It's the magnetic field inside the rotor, attracts micro fine iron dust and kaboom. When I was on the tools, one of the factories on my patch used to machine graphite components. No matter how well you sealed things, that dust always managed to find a way in to play with Mr Sparky. I use air tools for grinding iron. Apart from being short proof, they also stall rather than snatch and kick back if you get too enthusiastic with them.

[JohnD]

Perhaps you could insert the tool into a plastic sleeve?

But then it would need to be ventilated, which needs a compressor....................

 

I have one, initially to do heads with an air die grinder. Over the years, I've acquired several air tools, from a simple tyre inflation gauge, to a Desoutter air file (narrow belt sander) that can produce a gorgeous finish on welds. An air hacksaw is one that I've found very useful, while the air screwdriver I've never used in anger! But I'd not want to be without them. A compressor is a costly, but very useful ......... not 'tool', more 'resource' for the serious garagiste.

John

(PS. If you're thinking of getting one, mine has a direct drive from motor to compressor, rathar than a belt. Makes it compact, but also VERY NOISY!)

Edited May 1, 2012 by JohnD

[Wolfrace.]

You can drown out the worst of a direct drive compressors racket by putting it in a rockwell bat lined 'dog kennel'. The ends need to be open for ventilation but mount the ends also covered on the inside with Rockwell bats stood off - just make the top longer than the sides, so you keep a reasonable air gap. Sound wil not go 'rounded the corner' and Rockwool bats, (you can get proper acoustic grade if your feeling flash, but the ordinary ones from B&Q work pretty well), are good sound absorbers'. We used to use these in the workshops and they are surprisingly effective at taking the sting out of the racket.

[GHR630]

Rockwell bats?? Is this one for the RSPCA????

[Wolfrace.]

Damn predictive text!

[Nick Jones]

Got the head back on Friday with guides reamed and unleaded inserts fitted.

 

Just ground in the valves. The inlets sit a bit low in the seats (slightly bigger valves would be better) and you can see that there is quite a bit of unused seat area on the port side of the valve (the shiny bit).

 

So I popped it in the "lathe" (electric drill in a vice will suffice) ........

 

......... and back-cut the outer edge of the back of the valve until just short of the matt grey ring that indicates the ground in area.

 

Should give a little more flow, especially at low lift and is very quick to do.

 

cc'd the head and got 43.3 +/- 0.1cc. That gives approx 9:1 and I want 9.75:1 so I'm going to need to take about 0.85mm off. ONce that's done the head can be popped on and I'll have no excuse not to fit the engine..... deep joy.....

 

Nick

[GT6 Nick]

Which engine is this for? Not the one mucked about by the machine shop? Any news on that?

[Nick Jones]

No, this one is for the rebuilt PI engine (2.5) whereas the the one with cam bearing problems was intended as a replacement for the rather tired lump in the Vitesse (2.0), for use with the head I did a couple of years back. I have discussed the matter with the machine shop...... he says he will sort it......... so now I have to get the block back to him. As my work "territory" has now been extended to include where he hangs out, I just need need to find a victim customer/potential customer to go and see close by!

 

This head has gone back today for the final skim...... Need to sort it ASAP as the engine currently in the PI is getting sicker - making some nasty noises unless I'm really gentle with it......

 

Summer we are having, swapping the engines could be a problem as I don't have room to do it under cover - and it keeps pissing with rain

 

Nick

[Gerard]

Get yourself a nice big marquee Nick, it can always be useful later (barbecues, family do's and housing the offspring in the middle of winter ).

[GT]

I think you should bear this post very strongly in mind.

I took published results (in the public domain), so no-one can object to them being plotted out together.

 

These graphs are all measuring exactly the same things:-

 

STD heads for TR6 & GT6, and STD TC type valve heads for Spitfire.

 

I just used BRAND NEW OLD STOCK to be sure, and double checked with another very well known flow bench* that it was correct.

(My figures were within 1cfm between those benches!)

 

There are really huge differences in the results between different flow benches out there.

 

You can read a long thread from Vizard and Dave Baker about it, inc D Walker's archives, and the difficulties of calibration:-

 

http://web4.pistonheads.com/gassing/topic.asp?h=0&f=66&t=1105716&mid=0&i=20&nmt=when+dave+walker+talks+about+...&mid=0

 

What it boils down to is:-

 

P Burgess has a home made flow bench.

It clearly is reading massively & excessively optimistic at high flows. (20-25% plus!)

 

The SF110-120 series of flow benches are clearly reading quite non linearly in exactly the range 50-100cfm where it is most needed to be linear.

At lower lifts they start to correlate quite closely, and the error margin decreases.

 

I explain this by the fact the smaller benches are unable to run the 25" calibration standards because they aren't big enough, despite the fact I found almost no difference in the corrected results when testing at 10" or 25".

 

 

The thread above qualifies the error between just the small benches (SF120/1 and SF120/2) as being of the order of UP TO 8.5% (!)

This is enough to invalidate practically everything, unless you can establish proper bench marking and error rates, and again ilustrates the need for the proper calibration pack.

 

My tests used the much larger SF300-600 benches*.

I suspect APT in the USA is also using the large SF600 as his results are extremely similar to mine (really pessimistic).

 

It's up to you what you want to make of them, but note in passing, I DID do some brief back to back testing between the GT6 STD head and the TR6 STD head.

This is not something I normally waste time with, but it happened along the way.

 

The GT6 had a clear advantage of flow especially at higher valve lifts. (chamber depth related).

 

The TR6 head could be made to perform as well as the GT6 head by merely swopping out the original inlet valves for modern ones (as you can see on the graph).

 

The 219016 head had NO advantage whatsoever over the early TR6 head.

The only logic we could see was that Witor had deliberately doctored the figures for the early head, made them look worse, so that he could sell these later heads easier.

 

The back cut valves you decided to do, have a roughly equivalent effect of fitting a new more modern valve.

This advantage was easily quantifiable as being 2Bhp/cyl.

 

Here is the confirmation

 

Nothing is really quite what it seems.

[ncoll]

Firstly apologies to you nick for going of topic a bit

 

well folks

here we go again !!!! yes gareth your'e right and everyone else is wrong as usual. But just ask yourself who has had to apologise to me twice in as many weeks for getting things wrong. I was going to write a little article about the flowbench and how i calibrate it with the help of performance trends portflow analyser software programme, but no i will leave it up to you gareth to do it, although gareth i doubt if you own a flowbench, but then does that matter to you ? you seem to be the only person on the planet that has all the answers.

you know what they say " don't argue with idiots, they will drag you down to their level and beat you with experience" , so on that note i have more important things to do with my time, flowbenches and you gareth don't come into it.

 

neil

[GT]

WOW!

 

Mr C, This is getting silly.

All I have done is print YOUR OWN data, which you released yourself.

There's no rights or wrongs, it's all good old public domain stuff, so what's this "coughing over your cornflakes" stuff?

 

APT release their figures, You release your figures, Witor did that years ago (those were wrong btw).

 

I haven't released anything, but if you want me to, I will.

(you won't be disappointed).

 

People are (in the main) intelligent enough to draw their own conclusions.

 

What I added here proved that what Nick was doing on the valves would give him a measurable improvement in flow...NOT AT LOW LIFT, as he thought, but at maximum lift as you can see for yourself.

This is counter intuitive,but a fact of life,- because a fat lot happens on a Triumph head before 0.100" lift.

 

I did that test almost exactly 2yrs ago, testing 2 different modified valves in an absolutely standard port back to back against the Stanpart original one.

The result was quite impressive for such a simple test,- a bit like the 8V Golf GTI.

,- Swop out the heavy old 8mm valves, stick some decent 7mm ones, and you gain more than you'll get by porting it!

 

I understand this can create a massive hailstorm in your head.

It seems to be in fashion in northern England just now

Unfortunately there is no space for opinions or emotions on this, + as you say you have plenty else to do.

 

Here it's just all good old solid facts, which is what interests most people.

 

I spent a few hours & collated all the data.

 

As for ownership issues,- you know there's a SF600 for sale on EBAY just now for about 9500USD.

Almost brand new in fact you know in Charlotte, NC, United States.

 

http://www.ebay.com/itm/Superflow-600-Flowbench-/260876481820?pt=Motors_Automotive_Tools&vxp=mtr&hash=item3cbd73351c

 

If I were interested, I would buy it and ship it....it's only about double the price of the titanium stock I currently have just now, never mind all the piston forgings, and the other crap I have lying in my corridor.

 

Maybe I would make some sacrifices?

FFS forget the Jaguar rebuild this summer eh?

I have the measurement program, the neccessary com port, know where to buy a flow com, and even spare motors...

 

But ah no, I'm not at all interested in owning one, because the last people that benefited from all the work, couldn't even be bothered to pay their bills.

 

I'm trying to be nice, but you don't appear to be getting the message.

[Nick Jones]

Well, that is an interesting exercise Gareth, but I'm not quite sure what the argument is? I've always tended to the view that flow benches are a bit like rolling roads in that they are really best used as a tool for finding extra flow (or power if RR) "on the day" and comparisons with results obtained on other benches are always going need to be treated with caution.

 

Anyway, I don't have a flow bench (would like one though) and I'm not a racer aiming for the last few %, so my goals were to hopefully find a few more horses (and burn a bit less fuel) for the expenditure of modest amounts of money and effort. I've done this before (and the Vitesse goes well enough), but this time I had Neils tutorial to help out and this really helped me get a handle on what I should be trying do.

 

I think it's important to remember that the purpose of Neils thread was to give those amateurs inclined to have a go at their ports, some valuable guidance on how to do it so as to get some gain from the process. I think he succeeded in that, and for many of us, that's all we're really looking for. I can't justify a upper 3 figure, or even 4 figure sum for professional head work for a road car. In due course, this engine will get put back together, installed, tuned, run-in, and tuned again on my local RR. The RR figure won't be a useful comparison with anything, but at least I'll know whether I've got the best full load timing......

 

Nick

 

To extend the comparisons with rolling roads, my Vitesse showed 90 bhp at the wheels on a local rolling road (described by the owner as "conservative") - and that was all we could get - gained 1bhp in the session. I have had several people refusing to believe (on the grounds of being unable to stay with it) that it is only a 2.0, one of whom claims a rolling road figure of 170 bhp....... For damn sure I don't have anywhere near that - would be surprised to be within 45bhp. However, in that case he also had (unknown to him) the disadvantage of my car being driven by Duncan, my usual "co-driver" on events, who is a master of momentum management and can make that car absolutely fly - so the power was probably irrelevant anyway! Also probably irrelevant was that in the same car in the same few days we chased a 1.3 Twingo for about 60km through the Pyrenean foothills...... Twingo driver (bless him) knew the road really well and was trying hard (back wheel waving 8" off the ground in the bends gave that away) - we were just able to stay with him...... Don't think Twingos have 170 hp either.... Was a hell of a good drive

[Nick Jones]

Ah, typing too slow, interesting comment about the back-cut being more effective at higher lift Gareth. I wouldn't have expected that......

 

What also surprised me was just how little flow seems to go through the side of the valves at right angles to the port (as evidenced by the soot shown in the pics in my first post). I'd have expected more on the side by the exhaust valve than the side by the chamber wall......

 

Going off at a slight tangent, I've always wondered how comparable the actual flows are in a running engine (not steady state obviously) are with those obtained on a flow bench at steady state? I suppose it's irrelevant really provided best steady state flow also gives the best result on a running engine.

 

Nick

[GT]

I've always tended to the view that flow benches are a bit like rolling roads in that they are really best used as a tool for finding extra flow (or power if RR) "on the day" and comparisons with results .......to be treated with caution.

 

DON'T AGREE,

OK they're not ISO rated, but I spent a lot of time and money to get corrolation with another flow bench, so that my results from that one to the one I used for most work were not invalidated.

 

I don't agree it's impossible to compare, it's all a question of taking a known sample from one flow bench and testing it on the other one.....which of course people don't want to do....JUST IN CASE (of many things!)

 

Suprise, suprise on the PRO flow benches we were accurate to 1cfm.

You can see they (almost) are with APTFAST in the USA too.

 

Same was true on my Bosch rolling road.

I knew it was reading 94kw for Mr average TR5, 98kw for an good one and 104 for the 1 in a million (blue print standard).

What did that mean?

Is that really 128bhp at the wheels?? That can't be right.

 

Officially after checking an engine on a proper Schenk engine dyno, then sticking it into a chassis, we knew the transmission and all was losing us only 17bhp, grand max 23 on IRS, not some crazy 30% run down figure people want to believe.

 

Does that mean the engine making 141ps at the wheels (RR) was making 184bhp?

NO CHANCE.

My RR was wrongly calibrated.

The blueprinted engine was possibly making 158, possibly only 150 > real WHP was 132 or maybe only 125?, which are all very common error rates.

It didn't stop people at track days swearing the car was non-standard though.

 

3/

I think if you can't rely on SOMEONE or SOMETHING doing the job properly then you might as well not bother.

It's exactly the same for balancing cranks too.

 

Why do I still balance cranks at the same bloke year in year out?

Because he has SCHENK crank balancer.

 

Forget the rest....Jackson Bradley, all that crap....let him explain to you why Schenk is the BEST.

How come Lotus and Rover couldn't make K series engines to last 5 mins at high power?

'Cos they couldn't balance the crank right!

 

3/ I had Neils tutorial to help out and this really helped me get a handle on what I should be trying do.

 

I think it's important to remember that the purpose of Neils thread was to give those amateurs inclined to have a go at their ports, some valuable guidance on how to do it so as to get some gain from the process.

 

Completely agree with that,- gave me all sorts of handy hints to look for, especially reading between the lines & with lots of help from xxxx-S, near Silverstone, who doesn't take any nonsense.