Targeted application specific exhaust port design...

I am posting this to back up my assertions that the header design has a great deal to do with how an exhaust port flows with an actual primary pipe attached. I checked:

D0VE C filled with brass 230 max cfm 1.75" valve
Procomp raised very much like the TFS street. 251 max. 1.76 tuliped exhaust valve.
D3 with no floor fill but a bunch of port work. TFS cutter on seats. 1.8" exhaust valve. 214 cfm max.
Decent flowing P-51 research prototype. 1.76" tuliped valve 250 max cfm. Later P-51's max out at 262 cfm.
Out of the box SCJ casting.
Iron cj with typical mid 80's exhaust port work tweeked by RHP

Heads were flowed with 2" primary header pipe 4" straight off of the flange with one 90 degree bend vs the crites #8 primary for the 67 galaxie which abruptly turns down off of the flange and has a 50 degree bend.

Dove straight / crites..... P/C straight / crites....D3 straight / crites... P-51 straight / crites.. SCJ straight / crites... Iron cj straight / crites

.200" 116 / 114..............115 / 113..................107 / 106..........116 / 115...........102 / 101.....................104 / 103
.300" 156 / 150..............160 / 151..................145 / 145..........151 / 147...........139 / 136.....................142 / 140
.400" 184 / 176..............198 / 187..................176 / 174..........178 / 174...........164 / 160.....................169 / 165
.500" 205 / 192..............226 / 207..................201 / 196..........204 / 200...........184 / 182.....................189 / 184
.600" 217 / 204..............238 / 218..................213 / 209..........227 / 219...........197 / 192.....................200 / 194
.700" 224 / 209..............245 / 223..................214 / 211..........238 / 230...........207 / 199.....................207 / 200
.800" 230 / 213..............251 / 226..................214 / 210..........250 / 239...........211 / 204.....................210 / 204

By turning the flow column right off of the seat at the short turn in the D3 casting (this is the same casting piece i had dave mclain test and the numbers are posted elsewhere here) I can mitigate the losses as a % of total flow when the header primarys turn right off of the flange as almost all passenger car headers do. All i am doing is biasing the flow column lower in the port to utilise the normally dead area. I give up some high lift flow by shaping the port as i do but get much better flow at 400 through 600 lifts. The simple fact of the matter is this. BBF chassis headers and exhaust port design turn the exhaust flow column almost 180 degrees. I thought why not begin to turn the flow comumn with in the port where i as a porter can control it so that i can mitigate the losses normally associated with the crappy shock tower header designs we have to deal with.

This is why i do not recommend nor typically fill exhaust port floors in the iron heads unless the combo will utilise headers with some length off of the flange before turn down. Picture the typical TFS street, edelbrock, blue thunder or P/C filled raised exhaust port. High velocity and very efficient. Now picture that same port emptying into a primary flange that like most ford chassis headers turns abruptly off of the flange. What we have is high velocity flow activity at the roof of that that port crashing into the roof of the primary pipe which is already turning down. This hurts flow by a significant amount vs. flowing that same raised and filled port with a primary that goes for some distance before beginning to turn down. Also give consideration to the fact that the filled floor exhaust ports are emptying into a still oval header primary where the bottom 1/3 of the primary pipe is not seeing any port flow. Kind of like the floor of a regular bbf iron port. Little activity.

Dave McLains dyno results seem to back up the flow bench data... A header change on a BBC from turn down to several inches off the flange before turn down netted 30 HP in a 600 HP application.

Jon Kaase sent me a prototype of the P-51 cylinder head design to flow test and give feedback on years ago. I asked him if he had tried filling the exhaust port floors. He stated that he did in fact try it and found not a single HP...

When I dyno my engine combos I always try to utilise the chassis headers to be used with the combo vs using a pair of dyno headers which have ideal geometry off of the flange but could not be utilised in most car chassis combos. HP numbers are always conservative but i can count on them to be real as installed.

What i am really driving at here is that in talking with Dave McLain, Paul Kane, Blake Cartright etc... I feel that there may be some gains to be had by simply looking at the header primary and exhaust port as a whole and porting appropriately.

In order to further research this phenomenon i am going to take the above procomp exhaust port and start to lower the floor until i see some high lift flow losses. I will then retest this same port to see if i am able to increase the exhaust flow through the typical turn off the flange BBF exhaust header flange.

I am just beginning to wonder if mother Ford might have had a good reason to not fill that floor back in the day and if in fact it was a design necessitated by the shock tower cars that the engine was destined to go into to compete in the muscle car wars???

I am just beginning to wonder if mother Ford might have had a good reason to not fill that floor back in the day and if in fact it was a design necessitated by the shock tower cars that the engine was destined to go into to compete in the muscle car wars???[/quote]



Good info Scottie.

Thanks for the work.

Good read.
Am i correct in reading the P/C's exhaust flows as well or better than the P51 more or less until the crites is on it?

The P-51's Scottie had were prototypes, the later flow in the upper 260's . Probably can get more out of them , but I feel n/a engines that 255-260 cfm is plenty until you start spraying a butt load of NOS.

JMHO

I ran a 5.85 with a set of cast iron heads with the exhaust ports flowing barley 200cfm.

I recently sent a chunk of an R&D piece to Dave McLain (It is a D3 casting) to verify my work on his superflow bench. I see no reason what so ever why a head ported like this would not support as much power as the doves do... 2.2" intake and 1.8" exhaust.

.200" 143 / 112.5 W/O pipe
.300" 221 / 156
.400" 277 / 187
.500" 310 / 199
.600" 323 / 196
.700" 340 / 195

This should serve as verification to all of the naysayers that D0VE or D3 exhaust ports can flow 200 cfm WITH OUT PIPE. Look where that occurs. .500" lift. Considering that in an ideal world most exhaust evacuation is occuring by.300" lift during blow down this exhaust port example is not to shabby for old junk iron heads.

The whole point was to maximise flow rates in the mid lift ranges utilised by a majority of people and secondly...
To turn the exhaust flow column with in the port passage to accomodate the typical chassis headers that turn right off of the flange. This design shows a flow improvement with a primary header pipe bolted on both straight and abrupt turn down. The floor of this particular design port is quite a bit more active than a stock port's floor is...

When filled the port flowed great with a straight pipe at 234 cfm but suffered a bunch when an abrupt turn down was used down to 218 cfm.

D3 bare / W pipe / W turn sown pipe / D3 filled / W pipe / W turn down pipe

.200" 112 / 117 / 118 / 113 / 116 / 115
.300" 160 / 162 / 160 / 152 / 159 / 156
.400" 187 / 192 / 192 / 180 / 191 / 183
.500" 199 / 208 / 208 / 198 / 208 / 200
.600" 196 / 212 / 211 / 210 / 224 / 211
.700" 195 / 214 / 212 / 217 / 234 / 218

These numbers are from Dave McLains super flow bench. THANXX DAVE !!!

Filled floor port was noticably more turbulent with a turn down primary vs the std port with a turn down primary. Total flow in cfm .2 to .7" was better with the targeted port and turn down primary than the filled port with the turn down primary pipe.

Suffice to say it appears that in applications where the header primarys turn abruptly off of the flange a properly executed and targeted exhaust port enjoys a flow advantage vs the filled floor ports when the primary pipe is added to the equation.

Sticky this Bruno... Unless you are utilising an upswept header or one where the primarys are straight off of the flange for at least 3 to 4" Do not fill the exhaust port on a passenger car iron head or any head for that matter...

I guess it comes down to the question... How much exhaust flow is enough??? In speaking with Blake Cartright he was telling me that he switched out his really good TFS streets making over 800 HP with a really good pair of iron CJ's he ported having massive intake flow and 196 cfm exhaust flow. The combo's were remarkably close. General consensus seems to be that a 250 cfm exhaust port is plenty for any N/A BBF combo. Jon Kaase backed this up when he stated that in his 804 HP P-51 headed 521 test mule porting the exhaust from as delivered to 260 cfm and found not one HP.

More later...

Diggindeeper wrote:Good read.
Am i correct in reading the P/C's exhaust flows as well or better than the P51 more or less until the crites is on it?

Those procomp exhaust ports were highly worked with large valves and a tfs street cutter valve job. The P-51 numbers were early prototypes with a good cleanup but no port work.

The P-51's will go in to the 260 range routinely.

nice work Scott ,thank's for sharing!!
Frank

Frank Merkl wrote:nice work Scott ,thank's for sharing!!
Frank

X2

X2 on the info Scotty,
But respectfully,Id like to hear from the guys that have ran the filled floor setups also.
Maybe we can get some before/after results, along with what type of header setups they used with them.
I remember Nevs saying they had a pair and gained,but didnt elaborate.
Myself Im interested in the brazed floor route myself.
Thanks.
Al

Wow Scotty, that is some great info. After all the explaining it makes perfect sense!!

Its really cool of you to share this with us all!!

Bill

Scottie, I may have told you wrong about the Cast SCJ'S , they ran close on E.T. but were down on mph. It seems the week exhaust port hurt the hp in the upper rpm range but I think the good intake flow made that back up in the 5500 to 6500 range , above 6500 the car didn't seem to pull as hard.

Noticable seat of the pants feel. And the mph shows this. Of course the cam was not changed. One ground to crutch the lower exhaust numbers may have helped the mph. Probably a more narrow seperation angle would have helped also.

But bottom line is the old saying a Ford won't make power because of the weak exhaust flow seems to be B.S.. You only need enough to get the job done.

I have also found that if you do the normal exhaust hump removal and a decent port job on the exhaust ( 185-190 cfm, and don't touch the inake like most people do. A hydraulic cam in the 230 range at .050 with around .550 lift and a 107 -109 seperation runs best with very little split or none , (single pattern) . This is track tested not dyno.

With both intake and exaust ported , it seems like, from my dealings 6 - 8 degrees of split in a hydraulic cam works best , again in the 106 to 109 seperation range. depending on the @ .050 numbers your working with.

IDT-572 wrote:Scottie, I may have told you wrong about the Cast SCJ'S , they ran close on E.T. but were down on mph. It seems the week exhaust port hurt the hp in the upper rpm range but I think the good intake flow made that back up in the 5500 to 6500 range , above 6500 the car didn't seem to pull as hard.

Noticable seat of the pants feel. And the mph shows this. Of course the cam was not changed. One ground to crutch the lower exhaust numbers may have helped the mph. Probably a more narrow seperation angle would have helped also.

But bottom line is the old saying a Ford won't make power because of the weak exhaust flow seems to be B.S.. You only need enough to get the job done.

I have also found that if you do the normal exhaust hump removal and a decent port job on the exhaust ( 185-190 cfm, and don't touch the inake like most people do. A hydraulic cam in the 230 range at .050 with around .550 lift and a 107 -109 seperation runs best with very little split or none , (single pattern) . This is track tested not dyno.

With both intake and exaust ported , it seems like, from my dealings 6 - 8 degrees of split in a hydraulic cam works best , again in the 106 to 109 seperation range. depending on the @ .050 numbers your working with.

Most of the dyno work I have seen comparing the iron CJ's to aluminum heads with better exhaust ports has the iron CJ's about 25 hp short on the top side because of the exhaust port. I certainly agree that a tighter lobe sep is appropriate for the iron cj heads and even 10 degrees of additional exhaust timing. Its a matter of grinding a cam to suite the heads you are using...
The fact that those dinosaurs will support in excess of 800 HP is amazing as is the fact that you can make just shy of 700 hp with doves at street compression ratios and 730 plus with some squeeze...

--------------------------------------------------------------------------------

After gently dropping the floor of the filled and raised p/c exhaust port I found modest cfm flow gains throughout the lift curve. The port quieted a bit. As a percentage gain vs the original flow loss the increases were more significant at .500" and up. Given the pressure during blow down vs the 1 psi / 28" water a bench flows there could be some hp there.

P/C with pipe straight off of flange n / crites primary curved right off of flange / crites flow after port mods.

.200" 115 / 113 / 120 +7 cfm over flat filled floor.
.300" 160 / 151 / 153 +2
.400" 198 / 187 / 190 +3
.500" 226 / 207 / 212 +5
.600" 238 / 218 / 226 +8
.700" 245 / 223 / 234 +9
.750" 251 / 226 / 241 +15

Rolling the floor down carefully and gradually from .625" in and dropping .180" did not effect flow with the straight pipe attached except at very high lifts .750" + and helped with the crites primary pipe. As typically flowed on a pusher bench high lift flow will almost certainly suffer some with out pipe.

I think there is some merit here when and ONLY when utilising chassis headers that turn immidately off of the flange. Angling the header primary to match the port exit angle did not help flow to any significant degree.

In testing the different primary pipes in the crites set i found as expected that the greater the number of bends and or the tighter they were the greater the effects on flow.

Has anyone ever put a set of iron manifolds on a 600 hp motor and dynoed it? The lip of the manifold doesn't just turn, it is flat out for over an inch, almost two and the top goes up some. I wonder if that along with a good secondary, like an X pipe and 3" exhaust wouldn't do just as well as the Crites. to match too.

Another idea would be to bandsaw the mounting flange off of another set of manifolds and machine it all to give a good 2"-3" of flat floor before turning and mount the manifold to iron manifolds to that with a coat of gray silicone between and studs. You could sweep the flange up and machine the manifold to match it. Ford got so many things right on the 385 it makes you wonder if headers are even neeeded. Most of your gain is in the secondary anyway.

Just a thought. I don't have a dyno or a bandsaw or flat mill, but if someone did, it might be worth a try. The Crites hurt it by 6% on all 8 at 28" or 2 pKa. At real running pressure its closer to 600 pKa. And consider it goes up and down so the flow range is twice percycle. The real loss of all 8 cylinders at 6000 this is really 14%, close to the output of a whole cylinder. It makes me wonder if getting the floor shelf and secondary plumbing like an X pipe with 3" tubing from the manifold all the way right wouldn't gain you more than say 30" inches of 2" diameter of pipe would. It would simplify fitting for sure.

BillBallinger wrote:Has anyone ever put a set of iron manifolds on a 600 hp motor and dynoed it? The lip of the manifold doesn't just turn, it is flat out for over an inch, almost two and the top goes up some. I wonder if that along with a good secondary, like an X pipe and 3" exhaust wouldn't do just as well as the Crites. to match too.

Another idea would be to bandsaw the mounting flange off of another set of manifolds and machine it all to give a good 2"-3" of flat floor before turning and mount the manifold to iron manifolds to that with a coat of gray silicone between and studs. You could sweep the flange up and machine the manifold to match it. Ford got so many things right on the 385 it makes you wonder if headers are even neeeded. Most of your gain is in the secondary anyway.

Just a thought. I don't have a dyno or a bandsaw or flat mill, but if someone did, it might be worth a try. The Crites hurt it by 6% on all 8 at 28" or 2 pKa. At real running pressure its closer to 600 pKa. And consider it goes up and down so the flow range is twice percycle. The real loss of all 8 cylinders at 6000 this is really 14%, close to the output of a whole cylinder. It makes me wonder if getting the floor shelf and secondary plumbing like an X pipe with 3" tubing from the manifold all the way right wouldn't gain you more than say 30" inches of 2" diameter of pipe would. It would simplify fitting for sure.

Dave Mc wote me this morning and said the manifolds are definitely not good, wrong tree there

They drop too quick and narrow down in the just the wrong place. I am hoping to find a set of used 2" tube '72-'73 Torino headers that will fit a '65 Galaxie swap. Anyone have a set of wall hangers let me know or anything else for a '65 Galaxie swap. Thanks guys, please excuse my newbbieness. BTW, since I stuck my foot in my mouth good on that one... If I have to go with crites, is there anything I can do in the secondary to make up the difference? The frame is a '66, from the original rusting out, so I think the tranks crossmember just has to be flipped around and the ebrake stuff remounted to fit a C6. Thanks.