Cylinder Head Flow Benches (recent tuning threads)

[The Excession]

Given the recent contributions from our 'very own' (can I say that?) Pumaracing, and the appearance of Mr Vizard regarding porting cylinder heads, I'm utterly enthralled by the discussions and information that is being posted.

I'm no engine tuner, though I have rebuilt a couple of bike a car engines in my time, so with no disrespect to all the gentlemen involved...

It's like being aged 5 or 6 and sitting on Grandad's knee and him telling stories after a roast chicken dinner with Hovis bread where each slice was buttered and then all the butter was scraped off again ready for the next slice.

I would really like to know how you use and work with a flow bench.

I'm not looking for any trade secrets here, so to illustrate;

My background is in computer programming, and I remember a long long long time ago someone asking me 'How do you write HTML for a web site pages?'

I got all excited and started blabbing on about HTML, and the lad I was with stopped me and said 'No! You've not understood my question.... I can read all about HTML in a book, but how do you write and test it?'.

So I showed him Notepad, showed him some short cut keys for Internet Explorer and Windows and within a few days he was banging out HTML web pages.

Another time I insisted it was possible to write HTML without touching the computer's mouse and made that person work only using the keyboard. i.e. Change the document content, switch to the browser, view the changes, hop back to the document and change again - no mouse involed for copying, pasting, editing etc.

So, process over knowledge. I haven't got the knowledge (likely I never will), but I'm very interested in the process.

So my question is, how do you work day to day, hour to hour with a flow bench?

Obviously you hook up 'something' (a cylinder head), you make some measurements, then you might want to change something. Does that involve lugging a chunk of metal to the work bench and grinding stuff off and then testing and measuring it again?

If any of you would like to talk about computer models and fluid dynamic 'scenarios' even better, but what I'm really interested in is how you go from test 1 to results to engineering the input for test 2.

I'm quite fascinated by all of this!

[virgil]

Same here! as you say there's so much theory out there (in all walks of life) what we really need is a practical example.

I'm thiniing there's an awefull lot of thoery worked out in advance and the experience bit shows where to grind stuff off (or where to weld up) and reprofile.

I've been thinking of the best way to do this and the way I'm looking at is:

Take scrap casting and cut up to show profile OR take latex or silicone moulding of each port. So you have a good idea of the actual existing shape.

make a dummy (or copy) of the port in fibreglass, clay etc and use that on the flow bench to experiment with, save screing up your good head.

You can then try (using the fixed dutum points of the head in quesiton - head manifold faces, value andgle and position in the head) to create an 'ideal port' - I.e. a smooth tube leading the seat and test the flow. you now have an ideal (but impossibel to acheive) port to try and get as close to as possible. now morph the ideal to what's available (spare metal wizse) in the head and grind/fill to get as close to the ideal as you can.

Then measure it and find out you've just created a head flowing less than a bogoo std 4.6 head, wasted 3 weeks of evening sand scrapped a good tvr head! :-)

[MattYorke]

What I've found quite interesting is lots of things "seem like a good idea", but you rarely get a chance to try them all out, so it's a rare pleasure to read stuff from guys who basically, have.

Re flowbenches, all the stats I ever see give flow at various valve lifts, but referring to port velocity, can any flow benches measure either the velocity, or, I guess more completely (since for velocity to be valid, you'd need to measure it at many points) the momentum of the gas in the port.

I guess it should in theory be possible to work out, or measure, the resonant frequency of an intake or exhaust port (and relevant gubbins also bolted to it) and correlate this with the cam characteristics.

[Rwdfords]

I am not trying to send anyone away from this forum but there is a lot of useful flowbench information here:

http://www.turbosport.co.uk/showthread.php?138348-...

Read through all the posts and you will have a very good understanding how a flowbench works and how to build a Dtec orifice style flowbench, based on this proven accurate + repeatable design http://dtec.net.au/Flowbench%20Design%20Guide.htm

Regards
Jason

[Pumaracing]

The Excession said:

So my question is, how do you work day to day, hour to hour with a flow bench?

Obviously you hook up 'something' (a cylinder head), you make some measurements, then you might want to change something. Does that involve lugging a chunk of metal to the work bench and grinding stuff off and then testing and measuring it again?

Basically yes. You take flow readings every 50 thou of valve lift normally. You start with a test of the stock head and then you keep making changes to the port shape and size, valve seat profiles and chamber shape and test flow after every mod. It's very time consuming. A port can take an infinite variety of shapes so it's never possible to say you've found the optimum which is why flow development is still going on relentlessly with the Chevy V8 heads after more than 50 years and countless thousands of man hours of flow tests.

The important thing is to understand realistic flow targets i.e. the potential maximum flow from a given valve size in a port of a given downdraft angle so you know whether you're close to optimum or still a long way from it. The aim is to get the most flow from the smallest and therefore most efficient port size.

I suppose you could also say the ideal port shape is as straight as possible from manifold flange to valve seat so you remove metal to try and straighten out any bends that are in there and maximise the radius of any bends you can't fully straighten.

After a few years of flow testing you learn what is required to get most port shapes to flow well and then you can go straight to a fairly optimal size and shape without using the bench any more unless every last CFM is critical.

The hardest thing to get right by hand is the radius of the short side bend leading into the valve throat. Getting that smooth and of as large a radius as possible when most of the time you can't even see the end of the cutter you're using is more a matter of art and sculpture than physics. I still struggle with it. You have to imagine the shape you want in three dimensions, direct your hands to move the cutter in that shape and try not to make undercuts at the far end of each stroke as the cutter dwells before you pull it back.

With circular ports which most heads have I set the final size using old scrap valves with the heads machined down to the target size. I have a range of those in 1mm increments made on my lathe which I can push down a port and see where any tight spots are. Then I work at those spots until the target valve head size just fits right down the port from manifold flange to valve throat. If I need a custom size I make a new port probe on the lathe from another old valve. I can then guarantee that every port is the same size to within a few thou.

With rectangular ports it's more difficult and you have to do it by eye and by measurement with calipers.

The skills involved to do it right are vastly undervalued compared with things like law where solicitors charge £200 or more an hour for doing bugger all or computer programming or accountancy. I have backgrounds in all of those but getting cylinder heads right is the hardest of all and the one that makes you the least money. If I wanted to be well off I'd have stuck with accountancy which I originally trained in. However it's quite satisfying to beat every other tuning firm out there on the race track or the dyno and I'm not very money oriented so I guess I'm not that unhappy with my choice.

[The Excession]

Thanks Pumaracing, very interesting stuff.

I suppose my next question would be instead of working on an actuall head which costs money if you mess it up, is there a possibility to work on a kind of mock up cast say using plaster of paris?

My other thought would be the use of these new fangled 3D printers where the 'final' port profile could be produced in plastic and tested with minimum labour.

[Mr2Mike]

Pumaracing said:

It's very time consuming. A port can take an infinite variety of shapes so it's never possible to say you've found the optimum which is why flow development is still going on relentlessly with the Chevy V8 heads after more than 50 years and countless thousands of man hours of flow tests.

I remember some years back a process where an abrasive paste was pumped through the ports under very high pressure, which naturally eroded the material causing the most restriction. Is this still done, or did it not give effective results?

[varnish5000]

I think that CNC heads will become more affordable in the future.
With CAD models and the use of CFD in designing heads, testing will be quicker and cheaper.
The result will more than likely be better than hand produced ports using traditional testing methods as well.

[anonymous-user]

Mr2Mike said:

Pumaracing said:

It's very time consuming. A port can take an infinite variety of shapes so it's never possible to say you've found the optimum which is why flow development is still going on relentlessly with the Chevy V8 heads after more than 50 years and countless thousands of man hours of flow tests.

I remember some years back a process where an abrasive paste was pumped through the ports under very high pressure, which naturally eroded the material causing the most restriction. Is this still done, or did it not give effective results?

The Process is called "flow honing" and it is still used for port modifications. It's expensive however, so generally only used for prototype or race port finishing.

[Rwdfords]

Pumaracing said:

The important thing is to understand realistic flow targets i.e. the potential maximum flow from a given valve size in a port of a given downdraft angle so you know whether you're close to optimum or still a long way from it. The aim is to get the most flow from the smallest and therefore most efficient port size.

Dave, can I ask how does one work out what reasonable flow rate to aim for with a given port angle and valve size?

What would be your estimate for a 2.0 pinto injection inlet port with a 46mm valve @28" and at what lift?

If one was aiming for 220bhp+ from a 2.0 pinto what flow rate would you be aiming for?

We will be working on such an engine this year with 15mm lift roller cam, around 255 to 260* @.050" with lash, massive valve open area in the upper lift range compared to a flat pad cam and much shorter seat timing

Do you aim to have your ports max flow a little before peak valve lift to perk up the lower lift flow and fill the short turn bottom of port to get higher velocity/better cylinder filling? I was thinking something around 37 to 38mm maximum port diameter with 46mm valves

Regards
Jason

[Pumaracing]

Rwdfords said:

Dave, can I ask how does one work out what reasonable flow rate to aim for with a given port angle and valve size?

What would be your estimate for a 2.0 pinto injection inlet port with a 46mm valve @28" and at what lift?

If one was aiming for 220bhp+ from a 2.0 pinto what flow rate would you be aiming for?

We will be working on such an engine this year with 15mm lift roller cam, around 255 to 260* @.050" with lash, massive valve open area in the upper lift range compared to a flat pad cam and much shorter seat timing

Do you aim to have your ports max flow a little before peak valve lift to perk up the lower lift flow and fill the short turn bottom of port to get higher velocity/better cylinder filling? I was thinking something around 37 to 38mm maximum port diameter with 46mm valves

Regards
Jason

I work at 25" not 28". Only the septics work at 28" and that's only because Smokey Eunuch decided 50 years ago that the extra 3" of pressure drop was oh so vitally important and all the sheeples over there followed suit. A plague on the lot of them.

Anyway I've never done a filled port Pinto but never really needed to. I hit 194 CFM at 550 thou with a 45.7 mm valve after a day or so on the flowbench and that's plenty for well over 200 bhp. That agrees pretty exactly with DV's best ever of 188 cfm at the same lift from a 44.5 mm valve on a non-filled port as per his book. I'm sure there was a bit more to come but when other people's big valve heads that I tested were barely breaking 160 cfm it seemed like enough to be going on with.

I've done a couple of 200 bhp engines with only 44.5 mm valves and conventional cams so a 46 mm head with roller cam ought to make 220 bhp on 195 ish cfm.

Your port sizes are way out. A 36 mm straight tube should flow 217 cfm so the stock Pinto port is big enough for even filled port heads.

You other questions will have to wait for another day.

[Rwdfords]

Thank you for your detailed reply

Agreed on the 28" vs 25"

Those flow figure targets seem quite reasonable, I have managed 181cfm with a non flowbench ported injection head without filler but the inj inlet ports are quite good to start with (that engine made 190bp/157lb/ft 2.0 with a pretty mild cam), with a 46mm inlet hopefully we should be flowing 205 to 210cfm@25" and .550" lift or so with some port filler and a very nice port shape

I agree there are some really rubbish heads out there with big valves fitted and the ports just cleaned up and nothing else leaving the bowl areas very small in relation to the valve size and a sharp edge on the short side turn lol

Good to know we should have enough cfm to make some very decent hp when everything else is dialled in

I didn't realise 36mm would be enough for our needs, that would certainly improve cylinder filling, very interesting looking at DV's pinto power curve with filled ports, it makes a lot of sense how using the smallest port possible with enough high lift flow for the top end power you need, like some of these highly developed R1 and CBR race bike ports

http://www.racedevelopments.co.uk/flowbench/index....

http://www.racedevelopments.co.uk/cylinder_headwor...

From what you and DV have said about high velocity ports I am more than willing to try this

Do you aim for something similar with the exhaust ports or does the valve throat area act more as the venturi/high velocity point in exhaust ports? should the port be getting steadily larger after the throat area?

Regards
Jason

[Pumaracing]

Rwdfords said:

Do you aim for something similar with the exhaust ports or does the valve throat area act more as the venturi/high velocity point in exhaust ports? should the port be getting steadily larger after the throat area.

I fret very little about exhaust ports and don't usually even bother flow testing them. A 10% gain in inlet flow = 10% extra power but on an exhaust port it might only be 1% extra power unless the ratio of exhaust to inlet flow is abnormally low. Not worth getting worked up about. Aim for a discharge coefficient of 0.65 so make the port area about 65% of the valve area or a tad over - say a diameter of 81% to 83% of the valve diameter for a round port, bore the throat to 86% to 88% of the valve diameter, cut a nice seat per my previous recommendations in the valve seat thread and spend the rest of your time worrying about the inlet side.

I'd rather have a tight exhaust port with high gas speed and good anti-reversion properties as well as a decent mismatch to the exhaust manifold than taper the port out to try and gain pressure recovery. Aim for the biggest mismatch on the bottom side of the port to the exhaust manifold you can get i.e. open the port up on the top side at the manifold end for preference. That'll stop reversion down the bottom side of the port and back round the short side bend.

The stock 36mm exhaust valve with a properly modified port is just about big enough for even a 46mm inlet so I've never bothered using Group 1 exhaust valves unless they've already been fitted. In fact I rarely fit bigger exhaust valves on any race engine. On some engines I make them smaller. Look at the ratio of Chevy exhaust to inlet sizes for a better understanding of this. A 1.6" exhaust with a 2.1" inlet on a Chevy LS would be equivalent to only a 35mm exhaust valve with a 46mm Pinto inlet. I'm not saying that's an ideal ratio but they get away with it.

You do need to get the Pinto exhaust port lateral bend properly modified though. As stock it doesn't flow for st. Maybe 100 cfm or so. Take as much metal as you can off the inside of the bend (and some off the outside) and straighten it out as best you can. You'll get close to a waterway if you do this right. If you can get 140 cfm you'll be in the ballpark if you can be arsed to flowtest anything. I can't until someone else gets close enough to make me actually start trying hard.

One more advantage of the stock exhaust valve is it's less chamber shrouded than the 38.1mm Group 1 valve so you need to do less in the chamber to get it to flow right. In fact it's hard to get 38mm or 39mm valves to flow much more than the 36mm one. That doesn't stop everyone and his dog from thinking bigger must be better of course so they just cock up a potentially decent engine with big exhaust valves and crap port shapes which flow worse than a stock valve and a decent port shape. Humans eh? Loathe em or hate em they'll always cock stuff up given half a chance.

[Stan Weiss]

There are a number of calculated number that can be used to compare different heads. The DC (Discharge Coefficient) in different areas of the head is one that is popular. The flow per sq. inch of valve area is another. Here is an example using a 2.0-2.3 Duratec ported head. By the way does anyone have numbers for a 2.5l Duratec head?

Stan

Bore = 3.445     Stroke = 3.258            Rod Length = 5.758     RPM = 6500
Wrist Pin Offset = 0.0                     Number of - Intake Valves = 2 - Exhaust Valves = 2
Intake Valve Size = 1.378                  Exhaust Valve Size = 1.181
Intake Valve / Bore Ratio = 0.4            Exhaust Valve / Bore Ratio = 0.342816
Intake Valve Area = 2.98276 sq. in.        Exhaust Valve Area = 2.190885 sq. in.
Intake Valve Stem Size = 0.3415            Exhaust Valve Stem Size = 0.3415
Intake Valve Stem Area = 0.18319 sq. in.   Exhaust Valve Stem Area = 0.18319 sq. in.
Valve Lift at which the Valve Area and Window / Curtain Area are the SAME SIZE
At that point the velocity will be the same in both areas
Intake Valve Lift = 0.3445 Inches          Exhaust Valve Lift = 0.29525 Inches
Intake Centerline = 111.0                  User Selected DC - Discharge Coefficient = 0.5
Bowl CSA (0.91) Intake = 2.2868 sq. in.    Bowl CSA (0.91) Exhaust = 1.6311 sq. in.
Effective Bowl CSA = 1.23829               Effective Bowl CSA = 1.22023
Valve Lift at which the Bowl Area and Window / Curtain Area are the SAME SIZE
At that point the velocity will be the same in both areas
Intake Valve Lift = 0.52825 Inches         Exhaust Valve Lift = 0.43962 Inches


     Intake                Curtain       Effective        Bowl             Valve        CFM per Sq. In.   L/D
 Lift      CFM     fps       DC     Area    Area      fps       DC      fps       DC      Bowl    Valve  Ratio
 .1000   84.000  232.842   0.6647    .866    .576   88.157   0.2517   67.588   0.1929   36.732   28.162  .073
 .1500  124.000  229.146   0.6541   1.299    .850  130.136   0.3715   99.773   0.2848   54.223   41.572  .109
 .2000  163.000  225.912   0.6449   1.732   1.117  171.066   0.4883  131.154   0.3744   71.278   54.647  .145
 .2500  199.000  220.646   0.6299   2.165   1.363  208.848   0.5962  160.120   0.4571   87.020   66.717  .181
 .3000  227.000  209.743   0.5988   2.597   1.555  238.233   0.6801  182.650   0.5214   99.264   76.104  .218
 .3500  238.000  188.491   0.5381   3.030   1.631  249.778   0.7130  191.500   0.5467  104.074   79.792  .254
 .4000  246.000  170.474   0.4867   3.463   1.685  258.174   0.7370  197.937   0.5651  107.572   82.474  .290
 .4500  249.000  153.380   0.4379   3.896   1.706  261.322   0.7460  200.351   0.5719  108.884   83.480  .327
 .5000  253.000  140.260   0.4004   4.329   1.733  265.520   0.7580  203.570   0.5811  110.633   84.821  .363
Avg     198.111  196.766   0.5617   2.597   1.357  207.915   0.5935  159.405   0.4551   86.631   66.419


    Exhaust                Curtain       Effective        Bowl             Valve        CFM per Sq. In.   L/D
 Lift      CFM     fps       DC     Area    Area      fps       DC      fps       DC      Bowl    Valve  Ratio
 .1000   83.000  268.448   0.7663    .742    .569  122.127   0.3486   90.922   0.2596   50.886   37.884  .085
 .1500  121.000  260.901   0.7448   1.113    .829  178.041   0.5083  132.549   0.3784   74.184   55.229  .127
 .2000  144.000  232.870   0.6648   1.484    .987  211.884   0.6049  157.744   0.4503   88.285   65.727  .169
 .2500  157.000  203.115   0.5798   1.855   1.076  231.012   0.6595  171.985   0.4910   96.255   71.661  .212
 .3000  162.000  174.653   0.4986   2.226   1.110  238.369   0.6805  177.462   0.5066   99.321   73.943  .254
 .3500  168.000  155.247   0.4432   2.597   1.151  247.198   0.7057  184.035   0.5254  102.999   76.681  .296
 .4000  170.000  137.458   0.3924   2.968   1.165  250.141   0.7141  186.226   0.5316  104.225   77.594  .339
 .4500  175.000  125.779   0.3591   3.339   1.199  257.498   0.7351  191.703   0.5473  107.291   79.876  .381
 .5000  176.000  113.848   0.3250   3.710   1.206  258.969   0.7393  192.799   0.5504  107.904   80.333  .423
Avg     150.667  185.813   0.5304   2.226   1.032  221.693   0.6329  165.047   0.4712   92.372   68.770

Edited by Stan Weiss on Tuesday 17th January 18:47

Edited by Stan Weiss on Tuesday 17th January 18:50

[K87]

Not sure if it's just me puma but your website is being blocked as a trojan threat.

[Pumaracing]

K87 said:

Not sure if it's just me puma but your website is being blocked as a trojan threat.

Yeah there's a constant attack on it which I remedy every few days by replacing any corrupted files. Getting me right pissed off it is.

[Pumaracing]

Stan Weiss said:

There are a number of calculated number(s) sic that can be used to compare different heads.

God that's a horribly badly ported head. Only 0.58 DC on the inlet and even less, 0.55 on the exhaust? How the hell did they manage to make the exhaust flow worse than the inlet? A blind man's guide dog could do better than that. There's about 20 bhp to come from doing that head properly. Mind you that's about par for the course from the tuners out there.

[Stan Weiss]

This is a a couple of year old NHRA 500 ci Pro Stock head.

Stan

Bore = 4.72     Stroke = 3.57              Rod Length = 6.125     RPM = 10500
Wrist Pin Offset = 0.0                     Number of - Intake Valves = 1 - Exhaust Valves = 1
Intake Valve Size = 2.575                  Exhaust Valve Size = 1.865
Intake Valve / Bore Ratio = 0.545551       Exhaust Valve / Bore Ratio = 0.395127
Intake Valve Area = 5.207681 sq. in.       Exhaust Valve Area = 2.731792 sq. in.
Intake Valve Stem Size = 0.31              Exhaust Valve Stem Size = 0.3415
Intake Valve Stem Area = 0.075477 sq. in.  Exhaust Valve Stem Area = 0.091595 sq. in.
Valve Lift at which the Valve Area and Window / Curtain Area are the SAME SIZE
At that point the velocity will be the same in both areas
Intake Valve Lift = 0.64375 Inches         Exhaust Valve Lift = 0.46625 Inches
Intake Centerline = 111.0                  User Selected DC - Discharge Coefficient = 0.5
Bowl CSA (0.93) Intake = 4.4286 sq. in.    Bowl CSA (0.91) Exhaust = 2.1706 sq. in.
Effective Bowl CSA = 0.92217               Effective Bowl CSA = 0.89139
Valve Lift at which the Bowl Area and Window / Curtain Area are the SAME SIZE
At that point the velocity will be the same in both areas
Intake Valve Lift = 0.54745 Inches         Exhaust Valve Lift = 0.37047 Inches


     Intake                Curtain       Effective        Bowl             Valve        CFM per Sq. In.   L/D
 Lift      CFM     fps       DC     Area    Area      fps       DC      fps       DC      Bowl    Valve  Ratio
 .2000  175.000  259.593   0.7411   1.618   1.199   94.837   0.2707   80.650   0.2302   39.515   33.604  .078
 .3000  273.000  269.976   0.7707   2.427   1.870  147.946   0.4223  125.814   0.3592   61.644   52.423  .117
 .4000  360.000  267.009   0.7622   3.236   2.466  195.093   0.5569  165.909   0.4736   81.289   69.129  .155
 .5000  457.000  271.163   0.7741   4.045   3.131  247.660   0.7070  210.612   0.6012  103.192   87.755  .194
 .6000  530.000  262.065   0.7481   4.854   3.631  287.221   0.8199  244.255   0.6973  119.675  101.773  .233
 .7000  583.000  247.090   0.7054   5.663   3.994  315.943   0.9019  268.680   0.7670  131.643  111.950  .272
 .8000  614.000  227.700   0.6500   6.472   4.207  332.743   0.9499  282.967   0.8078  138.643  117.903  .311
 .9000  637.000  209.982   0.5994   7.281   4.364  345.207   0.9855  293.566   0.8380  143.836  122.319  .350
1.0000  644.000  191.060   0.5454   8.090   4.412  349.001   0.9963  296.792   0.8473  145.417  123.663  .388
Avg     474.778  245.071   0.6996   4.854   3.253  257.295   0.7345  218.805   0.6246  107.206   91.169


    Exhaust                Curtain       Effective        Bowl             Valve        CFM per Sq. In.   L/D
 Lift      CFM     fps       DC     Area    Area      fps       DC      fps       DC      Bowl    Valve  Ratio
 .2000  103.000  210.955   0.6022   1.172    .706  113.885   0.3251   90.490   0.2583   47.452   37.704  .107
 .3000  154.000  210.272   0.6003   1.758   1.055  170.275   0.4861  135.296   0.3862   70.948   56.373  .161
 .4000  200.000  204.811   0.5847   2.344   1.370  221.137   0.6313  175.709   0.5016   92.140   73.212  .214
 .5000  243.000  199.076   0.5683   2.930   1.665  268.681   0.7670  213.486   0.6094  111.951   88.953  .268
 .6000  286.000  195.253   0.5574   3.515   1.959  316.226   0.9027  251.264   0.7173  131.761  104.693  .322
 .7000  327.000  191.352   0.5463   4.101   2.240  361.559   1.0321  287.284   0.8201  150.649  119.702  .375
 .8000  344.000  176.137   0.5028   4.687   2.357  380.355   1.0858  302.219   0.8628  158.481  125.925  .429
 .9000  351.000  159.752   0.4560   5.273   2.405  388.095   1.1079  308.369   0.8803  161.706  128.487  .483
1.0000  360.000  147.464   0.4210   5.859   2.466  398.046   1.1363  316.276   0.9029  165.853  131.782  .536
Avg     263.111  188.341   0.5377   3.515   1.803  290.918   0.8305  231.155   0.6599  121.216   96.314

[Rwdfords]

Pumaracing said:

I fret very little about exhaust ports and don't usually even bother flow testing them. A 10% gain in inlet flow = 10% extra power but on an exhaust port it might only be 1% extra power unless the ratio of exhaust to inlet flow is abnormally low. Not worth getting worked up about. Aim for a discharge coefficient of 0.65 so make the port area about 65% of the valve area or a tad over - say a diameter of 81% to 83% of the valve diameter for a round port, bore the throat to 86% to 88% of the valve diameter, cut a nice seat per my previous recommendations in the valve seat thread and spend the rest of your time worrying about the inlet side.

I'd rather have a tight exhaust port with high gas speed and good anti-reversion properties as well as a decent mismatch to the exhaust manifold than taper the port out to try and gain pressure recovery. Aim for the biggest mismatch on the bottom side of the port to the exhaust manifold you can get i.e. open the port up on the top side at the manifold end for preference. That'll stop reversion down the bottom side of the port and back round the short side bend.

The stock 36mm exhaust valve with a properly modified port is just about big enough for even a 46mm inlet so I've never bothered using Group 1 exhaust valves unless they've already been fitted. In fact I rarely fit bigger exhaust valves on any race engine. On some engines I make them smaller. Look at the ratio of Chevy exhaust to inlet sizes for a better understanding of this. A 1.6" exhaust with a 2.1" inlet on a Chevy LS would be equivalent to only a 35mm exhaust valve with a 46mm Pinto inlet. I'm not saying that's an ideal ratio but they get away with it.

You do need to get the Pinto exhaust port lateral bend properly modified though. As stock it doesn't flow for st. Maybe 100 cfm or so. Take as much metal as you can off the inside of the bend (and some off the outside) and straighten it out as best you can. You'll get close to a waterway if you do this right. If you can get 140 cfm you'll be in the ballpark if you can be arsed to flowtest anything. I can't until someone else gets close enough to make me actually start trying hard.

One more advantage of the stock exhaust valve is it's less chamber shrouded than the 38.1mm Group 1 valve so you need to do less in the chamber to get it to flow right. In fact it's hard to get 38mm or 39mm valves to flow much more than the 36mm one. That doesn't stop everyone and his dog from thinking bigger must be better of course so they just cock up a potentially decent engine with big exhaust valves and crap port shapes which flow worse than a stock valve and a decent port shape. Humans eh? Loathe em or hate em they'll always cock stuff up given half a chance.

Hello Dave
Thank you once again for the detailed response

I agree the grp1 valves sizes are not anywhere near optimal, 46 - 37 seems about ideal, even 46.5 - 36.5 could work very well, you are right that is still well below the inlet to exhaust valve diameter ratio in some of the best US 2V engines using no extra exhaust duration or lift, it makes sense that the higher the SCR the smaller the exhaust can be made due to the different expansion ratio

I am for a very similar exhaust port shape to DV's, only porting in the high velocity areas, raised roof, widened towards the opposite side of the sharp turn, and making the turn as straight as possible without breaking through + widening the opposite side to minimise restriction, I think this would work much better than the huge rectangular ports I see in some race heads with poor velocity, the anti reversion ideas sound good, I usually open up the manifold flanges so that the roof of the port is level with the primary pipe and a step at the bottom and sides

One of the biggest problem areas with this engine seems to be the valve length/rocker geometry, finding valves the right length (short ending up with the with same geometry as std valves) and diameter is difficult to say the least

I have done some detailed testing with different length valves and no only do you loose a lot of lift with longer valves but it also alters the valve timing, it has the effect of tightening the LSA the more lift you loose, ending up with a lot mote overlap lift, one can tighten the LSA very easily with shorter valves by 3 or 4* and more like 6* longest vs shortest, so not only do you loose a load of lift and duration around peak lift but the LSA also gets really tight as low as 100* instead of the 106* that cam should have with short valves and the full quoted valve lift

It seems that the smaller the diameter valves used or the poorer the head flow the tighter the LSA needs to be to make best use out of the head, and the larger the valves / better the low lift flow (overlap flow) the wider the LSA needs to be to make best use out of the head

What sort of LSA's have you successfully used/what would you choose for a high output engine where a wide power band was needed for rallying?

Over the years we have used a lot of the Kent range with LSA's from 102 to 112* (going by the .050" lift points with proper lash), the best of which had 111* LSA, I am thinking 108 to 111* would work well with a short duration high lift profile with a suitable intake closing point, higher the duration the tighter the LSA or a higher SCR to get a good DCR

I think optimising the cam timing events is worth a lot of power from this engine along with increasing the lift as much a possible and lower the duration to a minimum, I would rather have an engine with huge torque than one with a high peak hp value and gutless out of corners, I see a lot of engines like that in competition, on the straights they never quite manage to make up what they lost out of corners, having to rev their engines to oblivion to make them go

Regards
Jason

[David Vizard]

Do you guys realize you have unleashed the porting monster in Puma Racing AKA Dave Baker. This guy is so fired up with hot tech on porting that I fear he is contributing to global warming!!!
(How did the move go Dave??_
DV