30 replies to this topic

[ScottNC]

There have been a couple of instances here in the states in the last couple years where with the help of designers in the U.K. crankshafts have been made for both NHRA Prostock and the NASCAR Cup series. In both instances improvements in power output (approximately 1-1.25%) above 9000RPM has been accomplished. I personally have have pursued such a crank but my employer at the time balked at the $50,000 price tag for two cranks (including design costs). It was explained to me that the power increase may not show conclusively on a conventional waterbrake dyno but would on an inertia or edie current dyno as well as on the racetrack as the improvement comes in part from management of the crankshafts own inertia. What are the design concerns/methods that would free-up power in an automotive 90 degree V8 as used in the two catagories listed? I have on good account that weight reduction through the use of high-end spec materials in the order of six or so pounds (NASCAR) was part of it. Also that bolt-on heavy metal counterweights were used.

[Greg Locock]

So, what are the absolute changes in power - 5-6 hp at 9000 rpm?

"It was explained to me that the power increase may not show conclusively on a conventional waterbrake dyno"

Why? an eddy current dyno is effectively the same, is it not? or do you do acceleration runs on an eddy current dyno, rather than fixed speeds (if so we are getting into BS territory)

"the improvement comes in part from management of the crankshafts own inertia. "

OK, where else?

Minimizing the main bearing drag is my best bet, which means dynamic modelling of the crank, so that each main bearing runs a minimum first and second order radial forcing. It would also be good to minimize crank bending so the mains run concentrically.

Reducing the crankshaft's inertia is not going to be very effective, except in first or second gear.

[ScottNC]

That the waterbrake dyno not showing an obvious power increase vs. an edie current came from Menard as did the enertia dyno comment, both during my request for a proposal from them. In NHRA Prostock the claimed increase was 10hp (waterbrake) at 9000 and 15 at 9500. I have no first hand power increase info on it's application in Cup other than to say that the individual involved in the NHRA application was involved in it's use in Cup first.

[ScottNC]

Exactly, what you wrote is what I'm after.........



"Minimizing the main bearing drag is my best bet, which means dynamic modelling of the crank, so that each main bearing runs a minimum first and second order radial forcing. It would also be good to minimize crank bending so the mains run concentrically."


I'm interested in being educated...........Can you explain how this is done? In the past observing bearing and crank friction points were giveaways as to where potential power might (was) be freed-up, ie; juggling counterweight locations to points opposite wear. I certainly don't know all I should about crankshaft balancing. It seems there must be a better way to go about it than to hang the greatest gob of weight on the extreme ends of a crank (as per current V8 manufacture) to "convienence" balance it at the expense of the rest of the shaft?


There appears to be a way to forecast the required necessities, both for mass and location without all my time consuming trial and (mostly) error?

[Greg Locock]

Above say 150 hz (9000 rpm, first order) the crank can no longer be considered to be rigid. (neither is the block, for that matter) So, the crankshaft designer optimises the counterweighting so that each web plane is balanced in its own right. One inefficient way is to divide the engine up into a long string of single cylinder engines, but more usually they just mess about with the couterweighting while trying to minimise the first order forcing at each bearing.

I've never done it, we paid ISVR in Southampton Uni to do that for us. $50k sounds cheap

[ScottNC]

Greg, from a design standpoint is this something strickly requiring computer modeling like FEA then? Empirical wisdom must play a part in it at some point, certainly the maths and/or programing was based on past cut-and-try results. Maybe once you've solved for one 90 degree V8 you've solved the key elements for all of them? I had thought to equip a conventional crankshaft balancer with transducers on all main journals as oppossed to just the two ends as a poor mans attempt to see what was happening. I suppose because of the low speed it may not tell you much....

[Greg Locock]

I'd do it using FEA, but you could put strain gauges onto each main bearing (or pressure transducers) and do it experimentally. Probably need to do both in reality.

As you say, you need to run at the right speed, there's no easy way round it.

[McGuire]

The typical eddy current dyno has significant inertia. If they say the power gain will only appear on an eddy current or inertia dyno, and not on a water turbine or AC dyno, what they are telling you is the power gain is mainly (or only) evident when the crankshaft is accelerating.

This makes perfect sense as one can see how the crankshaft would behave differently under acceleration than when relatively stabilized at a particular rpm. I'm not saying there's nothing to these approaches, only that the potential gains are out on the far edge of what is measurable.

A 1.0-1.25% gain is some mighty fine whittling. Many if not most race engine programs do not operate at that level of standard repeatability, especially if we are talking about NASCAR or NHRA pushrod engines. Lee White of Toyota tells me they can now validate all their NASCAR engines to within .5% of one another. By recent standards for this type of engine that is damn near metaphysical, so I don't blame him for bragging. In most of the series that run "identical" sealed spec engines, the supplier feels he is doing pretty good if he can gaurantee 1.5% from best to worst.

[knighty]

OK boys and girls, I have been visiting this forum for about 6 months now, and this thread has now prompted me to spill the beans - no big secrets if I'm honest!...... my information is based on the following facts.

I design engines for a living
I used to work in the TWR race engines department for 6 months prior to TWR going bankrupt………

When at TWR I saw a 90 degree cross-plane crankshaft for a NASCAR team, if I remember correctly it was Joe Gibbs racing........it had low pressure oil drillings, done on a 5-axis mill - very trick!

Now forget all this bull about CAE, dynos, counterweights and bearing drag, this is how a power gain comes from a high speed crankshaft……

Are you guys familiar with low pressure oil drillings?.........now the common method of doing the oil drillings is to do a single compound angle drilling through the main bearing, right up to the pin journal.

Way back when, in the days of the DFV – I think….. some bright spark, had the idea of doing the oil drillings so that the oil flow didn’t have to travel across the main crank journal, therefore the oil flow didn’t need to fight the high centrifugal force in order to travel from one side of the main journal to the other………they basically did a very short U-drilling, by drilling like so:-

Radially - Into the main journal about 10mm deep only
Axially – from the nose or flywheel end, to under the pin journal region
Radially – into the pin journal, through to the axial drilling above

Then block up the un-required openings with some grub screws………heypresto, low pressure oil drillings. I’m told that say a race engine needs 60psi of oil pressure at max RPM to work properly………with low pressure oil drillings, you could reduce the required oil pressure by about 20psi or so……….then you reclaim a few bhp due to reduced oil pump drag on the crankshaft

BUT

This was easy to do on a 180degree flat-plane crank……..but with the advent of 5-axis CNC machine stations……..with a bit of jiggery pokery, it can also be done on a 90degree cross plane crank……..TWR made two for Joe Gibbs racing, I,m told they worked very well, but I don’t know any power figures for you, it was a recommendation from Geoff Goddard, apparently the NASCAR boys had never seen this modification before, which I must admit I was quite surprised, as its been published in several “engine books”

Hope the above is of use to you

Do a web search on “low pressure oil drillings” and I’m sure you will find out more

[J. Edlund]

Originally posted by knighty
OK boys and girls, I have been visiting this forum for about 6 months now, and this thread has now prompted me to spill the beans - no big secrets if I'm honest!...... my information is based on the following facts.

I design engines for a living
I used to work in the TWR race engines department for 6 months prior to TWR going bankrupt………

When at TWR I saw a 90 degree cross-plane crankshaft for a NASCAR team, if I remember correctly it was Joe Gibbs racing........it had low pressure oil drillings, done on a 5-axis mill - very trick!

Now forget all this bull about CAE, dynos, counterweights and bearing drag, this is how a power gain comes from a high speed crankshaft……

Are you guys familiar with low pressure oil drillings?.........now the common method of doing the oil drillings is to do a single compound angle drilling through the main bearing, right up to the pin journal.

Way back when, in the days of the DFV – I think….. some bright spark, had the idea of doing the oil drillings so that the oil flow didn’t have to travel across the main crank journal, therefore the oil flow didn’t need to fight the high centrifugal force in order to travel from one side of the main journal to the other………they basically did a very short U-drilling, by drilling like so:-

Radially - Into the main journal about 10mm deep only
Axially – from the nose or flywheel end, to under the pin journal region
Radially – into the pin journal, through to the axial drilling above

Then block up the un-required openings with some grub screws………heypresto, low pressure oil drillings. I’m told that say a race engine needs 60psi of oil pressure at max RPM to work properly………with low pressure oil drillings, you could reduce the required oil pressure by about 20psi or so……….then you reclaim a few bhp due to reduced oil pump drag on the crankshaft

BUT

This was easy to do on a 180degree flat-plane crank……..but with the advent of 5-axis CNC machine stations……..with a bit of jiggery pokery, it can also be done on a 90degree cross plane crank……..TWR made two for Joe Gibbs racing, I,m told they worked very well, but I don’t know any power figures for you, it was a recommendation from Geoff Goddard, apparently the NASCAR boys had never seen this modification before, which I must admit I was quite surprised, as its been published in several “engine books”

Hope the above is of use to you

Do a web search on “low pressure oil drillings” and I’m sure you will find out more

Seems that it was used on the BRM engines before the DFV.

[desmo]

Not exactly on point, but have a look at Figure 3 on page 4 of SAE Paper 983036- Comparison Between V12 & W12 F1 Engines , presumably a Ferrari technique for F1 crank oil drillings.

[Greg Locock]

It's not bull, it is fact. Optimising the bearing loads is a fundamental part of crankshaft design.

[ScottNC]

Knighty; Gibbs is spot on. I know who the NHRA Prostock team is as well. Menards bought-out your former employeer according to the rumor in the sandbox?

McGuire; With current dyno systems in use in NASCAR, 1% repeatability is possible to achieve. Heck, on restrictor plate engines they worry to death over one horsepower or 1/4%. The single (in my opinion) biggest improvement is it is now common to have atmospherically (baro, inlet temp, vapor pressure) controlled engine inlet air instead of fighting a correction factor. It does not seem that any of the current correction factors in use are allways "accurate", that is, produces results in parrallell with engine responce to a spectrum of weather conditions. Further, it seem that for example, a NASCAR Cup engine would need a different correction for a given weather situation than would a Prostock drag engine. Ask anyone worth their salt at operating a dyno how hard it is to do an A-B-A engine test if you baseline in the morning and finish in the afternoon...without controlled inlet air. With it as long as the engine itself repeats a good operator can get quite good results, you may even go so far as to use only the observed numbers for testing. I will admit that with the bigger power numbers, like the 500 inch drag stuff, a small boo-boo in the math can wind-up being significant. There repeatability is an issue every day.

Unless you mean repeatability from engine to engine in which case ignore the above. I went back and reread your post and fear I have rambled needlessly. Since I went to the trouble of typing it up though (not a skill I have aquired) I'll leave it for what it's worth.

[ScottNC]

Knighty, would you mind if I contacted you by email? Thanks.

[McGuire]

Originally posted by ScottNC
Unless you mean repeatability from engine to engine in which case ignore the above. I went back and reread your post and fear I have rambled needlessly. Since I went to the trouble of typing it up though (not a skill I have aquired) I'll leave it for what it's worth.

yep, engine-to-engine is what I meant. Dynos were part of my job description for years, including NASCAR stuff....half to one percent repeatability for the installation itself is very do-able, but you have to keep on it like a hawk, just as you say. Many dyno operators don't, so when they see a one or two percent gain they have no idea what they are looking at.

You shouldn't need a different weather correction for different engines... they all breathe the same air eh. There are other corrections that may enter in, however, depending on the dyno installation & due to the differences in output scale. For engines that process very large relative volumes of air, Top Fuel for example, relative vs. absolute humidity becomes an issue. (You can't really dyno these monsters so this is more of a tuning issue.) So these guys use a measurement they borrowed from the commercial HVAC industry and track the specific air moisture content in grains per lb.

The current SAE standard correction factor (J1349) is to 77 F, 29.235 hg, 0' altitude, 0% relative humidity. However, much of the speed & performance industry uses the FAA/ICAO/ISA standard, which is 59 F, 29.92 hg. Free horsepower, with just a pencil and paper. Never buy a race engine in Denver. A good part of its output is correction factor. In F1, they often take the smart approach and do the best they can to mirror the actual running conditions to minimize the correction skew. Even down to piping a duct from the wind tunnel down to the dyno room for airbox pressure.

[McGuire]

Cosworth did not invent the "Low pressure crankshaft." It has been around since jesus was a corporal. Personally, I would amazed to see an old-fashioned cross-drilled crank in any recent NASCAR engine. "Cross-drilled crank" is another one of those hotrod buzzwords that in a real high-output, high-speed engine is in reality a bad idea. Unless you flood the crank with oil pressure, at high rpm it wipes out the rod bearings and windage management is impossible.

[ScottNC]

At present it is the "in" thing for a Cup team to engage engineering talent from F1 to help them. Not so 20 years ago. PMS or Pontiac Motorsports (back when there was such a thing) solicited help from Cosworth in the mid to late 80's. The results were pistons, pins and connecting rods that were conciderably lighter than was currently the trend. Most teams only tried said parts to satisfy Pontiac, convinced as they were that the stuff could never last. Can you say "self fufilling prophecy"? Anyway, along about this time I do remember crankshafts with odd oil hole drillings. As I remember (because it was a pain to de-burr these oil holes) the rod journal hole angled toward the main much as it does today but it intersected the hole from the main (which also angled in to meet the rod drilling) forming a sort of shallow "V" when viewed from the side. The hole from the main did not extend into the crank as far from a radial standpoint as was convention with a cross drilled or contemporarily drilled crank. As memory serves me there was an issue with these cranks because of the circumferal location of either the hole in the main journal or (more likely) the hole in the rod caused a timing issue with the oil supply. In retrospect this may have been an early attempt at low pressure oil holes on the Cup scene?

Having "Googled" myself into a stupor without success, is there a source for what is entailed in determining counterweight positioning beyond what has been elsewhere disscussed balance issues? Universally the typical V8 in the States has quite large first and last counterweights with most often (in the interest of overall weight) no center counterweights at all. When balancing an assembly the end weights are usually the only ones adjusted. Typically, wear is seen on the numbers two and four main bearings suggesting a bit of teeter-toter type goings on about number three main. Further, when running minimum piston deck clearance it is usual to see the center four pistons make incidental contact with the head before the four corner pistons. Limited satisfactory results have come from intuitively moving some of the inboard counterweights about slightly. This has been strickly "cut and try" methodology at best.

[McGuire]

Originally posted by ScottNC
Having "Googled" myself into a stupor without success, is there a source for what is entailed in determining counterweight positioning beyond what has been elsewhere disscussed balance issues? Universally the typical V8 in the States has quite large first and last counterweights with most often (in the interest of overall weight) no center counterweights at all. When balancing an assembly the end weights are usually the only ones adjusted. Typically, wear is seen on the numbers two and four main bearings suggesting a bit of teeter-toter type goings on about number three main. Further, when running minimum piston deck clearance it is usual to see the center four pistons make incidental contact with the head before the four corner pistons. Limited satisfactory results have come from intuitively moving some of the inboard counterweights about slightly. This has been strickly "cut and try" methodology at best.

What you want are textbooks produced before 1960... a good jumping-off point might be Automotive Engine Design by PM Heldt, 1933. Balancing & vibration is another of those areas that has taken a large step sideways in contemporary times that is not really relevant to ultimate-output racing engines or for lack of a better term, "advanced design."

Believe it or not, in recent times, if they have a troublesome crank vibration period in a particular production car application, they may just bend the crank or the block to shift it out of the critical range. The engine is now "smooth" as far as the passengers know, at the expense of friction hp (among other things, some of which you describe).

On the other hand, in earlier times there was a greater emphasis on classical solutions because that's all they knew. They didn't have computers and stuff to tell them how much they could chisel and weasel and still get away with it. They were forced to address problems straight-up and on the level.

For example, the old Y-block OHV Ford introduced in 1954 used a crankshaft with eight counterweights. The crank also had deep overlap, with 2.5" mains and 2.18" rod journals. (And of course the Y-block name signifies the full-skirted block, whch contrary to opinion was not to support the crank and main webs (they didn't touch) but the bell housing, so the driveline would not bend or twist the block and crank. That has traditionally been a problem with the Chevy, btw: the block only supports the bell housing through just over 180 degrees.) These are all important considerations in reducing friction hp. But on the other hand, the Y-block weighed over 650 lbs... Looks like I am rambling now...

The elegance & overkill approach to engineering has been rendered obsolete. Today the focus on that engine would be figuring out something cheaper to manufacture (and lighter) that was still halfway acceptable... then tuning up some big rubber mounts for it. But fortunately the books are still around that describe the core issues and direct solutions in plain terms.

[soubriquet]

Originally posted by McGuire

... Looks like I am rambling now...

Indeed. Why was it that excluding the Kent, Ford were unable to produce a decent engine in Europe until the current generation?

[McGuire]

Originally posted by soubriquet


Indeed. Why was it that excluding the Kent, Ford were unable to produce a decent engine in Europe until the current generation?

I never said that engines of the past were in general superior to those of today. Perhaps you are just in an argumentative mood.

[soubriquet]

Sorry, not trying to be argumentative at all! It is a genuine question.

I'm not familiar with the large engines, but I know the E93A (side valve, run in bearings), the 100E (side valve, shell bearings) and 105E (Kent) progression. The Kent was lovely, smooth and sweet, and it would rev. Then came the Pinto, a dog, and the CVH was rough as guts.

[knighty]

OK guys I stand corrected........jesus designed low pressure oil drillings when he was a corpral at BRM .......ref gaining power from counterweights......hmmmm..........I have designed several race cranks in my time.........one of particular note was a very "lightweight" 8-counterweight crank used in the ford Puma Super 1600 rally engines.........Farndons were pushed by Ford motorsport in Boreham to design the lightest crank possible, in the interest of keep ing the whole car weight low.......I think it weightd abour 9.5Kg........what a big big mistake .........ford went and homologated the crank - idiots........then it was found the crank counterweights were so light it was destroying the highly loaded centre main bearing.......requireing a mega expensive special bearing.......I designed a new 4 counterweight crank with about 45 Mpa bearing pressure for the centre main bearing, it weighed about 11Kg I think, it was right on the limit of the minimum allowed weight by the S1600 rules at the time........ and heypresto, the standard bearings worked quite happily with proper counterweight design..........but no power gain was found.........I would agree theres something in it.........but only a very small gain.........optimising bearing pressure is more an issue for bearing life rather than chasing a power gain in my opinion..........as they say - to finish first, first you gotta finish ;)

[knighty]

Scott - I have sent you my e-mail address via a personal post to you.

[McGuire]

Originally posted by soubriquet
Sorry, not trying to be argumentative at all! It is a genuine question.

I'm not familiar with the large engines, but I know the E93A (side valve, run in bearings), the 100E (side valve, shell bearings) and 105E (Kent) progression. The Kent was lovely, smooth and sweet, and it would rev. Then came the Pinto, a dog, and the CVH was rough as guts.

Exactly. Around that time, one of Ford's favorite ideas on that side of the puddle was the 60 degree V4. If they were serious about noise/vibration, they were starting by far on the wrong foot and they knew it. But they had become confident enough about controlling and masking these issues to go ahead anyway.

Same for the 90 degree Buick V6 and the half-a-V8 Pontiac slant four over here. GM was so confident (arrogant?) about their ability to turn sow's ears into something marketable they went ahead. If you look at these engines they have just one motivation vs. traditional "classical" designs: design, tooling and manufacturing costs. And bear in mind that at that time General Motors was the largest and most profitable corporation in the world, so it's not like they were working their way out of poverty.

I would never argue that today's engines are not the best ever. No disputing it. But change, even when clearly for the better, always contains elements of both good and bad...even in the most postive changes something worthwhile often can be left behind. There are lessons in the past, ideas and approaches from which we can profit here and now.

Who ever thought that F1 engines would discard the absolutely optimal solution, the bucket-type cam follower, and take a huge step "backward" to the rocker-arm, finger-type follower? But it happened, across the board. "Progess" is an interesting process, no?

[hydra]

The problem with buckets the way I see it is that they don't allow you to go really extreme with valve lift (there's only so much lift you can grind on a cam lobe) whereas finger followers allow you to attain the 16mm or so that modern F1 engines run. Plus the lobe itself can be made more aggressive, which is always a good thing. Plus with pneumatic valve actuation, valve control isn't as much of a problem as it was in the old days, so I guess that's why the went with finger followers...

[McGuire]

Originally posted by hydra
The problem with buckets the way I see it is that they don't allow you to go really extreme with valve lift (there's only so much lift you can grind on a cam lobe) whereas finger followers allow you to attain the 16mm or so that modern F1 engines run.

Right. As the issues and goals evolve, so must the solutions. The approaches of the past are not necessarily mistakes. They ceased to fit the puzzle at one point so they were set aside. As the puzzle continues to evolve we can go back and pick them up again.

[GeorgeTheCar]

As the puzzle continues to evolve we can go back and pick them up again.

Proving the need to understand history once again.

How soon we forget!

[Greg Locock]

"60 degree V4. If they were serious about noise/vibration,"

That would appear to be a mighty big IF

I know at least two of the big 3 developed cars that were killed because of insurmountable noise and vibration issues. At a previous employer we were asked to help sort them out within their program constraints, and, oddly, could not alter the laws of physics.

90 degree V6s, 60 degree V4s and even to an extent I3s require a lot of care and feeding in their mounting arrangements if they are to produce acceptable outcomes. I've even know I4s get delayed for a year while powertrain mounting is sorted. One program I worked on had 3 different mounting strategies built into their prototypes.

Even RWD cars with sensible engines are not immune, I worked on one car for two years in the Vehicle Semi-Anechoic Chamber that had the back of the gearbox supported on a test stand, as they could not develop a satisfactory gearbox mount. They solved it eventually, but it wasn't pretty.

[desmo]

Re finger follower vs direct acting digression: The following was posted here by someone who knows whereof he speaks,

"All F1 engines running today are of the finger following type, this allows for better cam profiles, as a direct acting cam on a bucket will have problems of acceleration up cam cheek , and tends to give severe wear and friction problems, plus needing very big buckets, thus increasing moving weight ."

[McGuire]

Originally posted by desmo
Re finger follower vs direct acting digression: The following was posted here by someone who knows whereof he speaks,

"All F1 engines running today are of the finger following type, this allows for better cam profiles, as a direct acting cam on a bucket will have problems of acceleration up cam cheek , and tends to give severe wear and friction problems, plus needing very big buckets, thus increasing moving weight ."

No doubt as to why the finger follower has reemerged. What is interesting to me: a configuration seen as retrograde or obsolete is now once again at the leading edge, while the setup regarded as optimum for so long has been pushed back out of the picture. The DOHC with bucket followers was THE classic layout for how many decades?

[McGuire]

Originally posted by Greg Locock
[B90 degree V6s, 60 degree V4s and even to an extent I3s require a lot of care and feeding in their mounting arrangements if they are to produce acceptable outcomes. I've even know I4s get delayed for a year while powertrain mounting is sorted. One program I worked on had 3 different mounting strategies built into their prototypes.

[/B]

Yep, I4's can be tough, especially the big ones.