V6 Strong modification

[FMG_V6_btb]

Maybe some of you would be interested in my project...

I have planned to modify deeply a 3.0 V6 24 Valve engine, to obtain a reliable 450-475 HP engine. The engine will be twin turbo, and the cylinder capacity maximized to be around 3.8 litres.

Since I have seen in the net many V6 engines with crank-rod failures (as mine), to build a reliable engine, means to solved this problem, and the better way (IMO) is to design a new crankshaft with increased stroke and rod journal diameter and width (lower bearing surface pressure):

Stroke: 72.6 mm (78 mm GTA) -> 82 mm
Main journal: 60 mm -> 62 mm
Rod journal: 52 mm -> 56 mm
Main-rod journal covering: 17 mm -> 18 mm
Piston diameter: 93 mm -> 99-100 mm -> 3787-3864 cc
Reinforced bottom end (maybe internal bedplate)

More features:

Forged aluminium pistons (around 9:1 CR)
Forged aluminium liners with nickasil coat
Forged steel rods
Dry sump lubricating system with one scavenge pump and two pressure pumps (one for internal rotating parts, and other for turbos and piston jets)
Two catalyst (engine must be homologated)
Individual throttle bodies with E-gas.
MBE engine management system with 2 pre-catalyst wide band lambda sensors and 2 post-catalyst jump type lambda sensors.
….And many more things.

I have machinery to do almost everything, I have simulation software for stress analysis and for engine cycle simulation (engine thermo & fluid dynamics), and the most important, engine test bench for until 800 HP engines. The only thing I have not enough is time…

Volunteers in collaborate with information, pictures, specifications, dimensions, measurements, or parts for experimentation, are welcome, we can share information and results.

Thanks.

[slyalfa]

was the rod fail with a stock rod or the H-beam rods?
if it was the stock rod I would say just go to the THS H-beam as it would cut a lot of cost. the set is around $900 and use std crank etc.
the stock rods seem to crack at the drilled oil hole.

[FMG_V6_btb]

No, the rod doesn't broke, the origin of the problem probably was a poor lubricating conditions in the bearing that allow to a heavy crank and rod wear.

If the engine power is going to be two times the actual at the same engine speed, that means the engine torque will be two times the actual, and this means the force in the rod will be two times the actual, I think the original crank and rod weren't designed to work in that conditions. To have a RELIABLE engine mean to design a new crank-rod-piston system taking into account the new conditions.

Thanks

[giuliettaevo]

You should speak to Barry H. He has build engines with these outputs before. You can also find him at ALFAGTV6.COM ~ Index

With a decent set of conrods and pistons i'm sure you an keep it together...

[slyalfa]

hmm I would think block metal fatigue would give in long before the crank.

I do think tieing the bottom togetter will help a lot. been thinking of that my self.

just making a crank and all custom parts seem like it would break the bank.
how were you going to make the crank?
as far as I know a billet type crank is weaker. but if the fail point is just the bearing a weaker crank might not matter.

it might be easer to upgrade the oiling on a stock crank? just thinking out loud. I would have to look at one (have one at home) but see if there can be improvements made to increase the oiling to the rods.
even if you do make a custom crank the improvements could be incorporated.

[oldschooleuros]

I would be more scared of increasing bore and stroke to almost a litre over factory, with boost added, that could be very disasterous.

[FMG_V6_btb]

No problem to increase bore and stroke if there is enough room to do it, and the parts are stiff enough also.

About room, many engines are using 101.5 mm pistons, and for a 82 mm stroke crank with new rods... that's something to check.

About the power, 450-475 HP for a 3.8 Liter engine means only 118-125 HP/Liter, which is not too much for a turbo engine.

Paco.

[FMG_V6_btb]

I’m going to start some calculations and simulations, first of the original engine (164 V6 24V QV) based on the following parameters:

Geometric inputs:
-Bore: 93 mm
-Stroke: 72.6 mm
-Conrod length: 131.1 mm
-Compression ratio: 10:1
Results:
-Combustion chamber volume: 54.796 cc
-Mean piston speed at 9000 rpm: 21.8 m/s (reasonable limit: 21-22 m/s)
-Mean piston speed at 6300 rpm: 15.2 m/s
-Max. connecting rod angle (inclination): 16.07 º
-Max. positive piston acceleration (at TDC) @ 6300 rpm: 20174 m/s2
-Max. negative piston acceleration (at BDC) @ 6300 rpm: 11425 m/s2
-Effective Mean Pressure at max. power (231 HP@6300 rpm): 10.94 bar
-Peak in-cylinder pressure at max. power: 92.6 bar at 12º ATDC (when tuned)
-Trapping efficiency at max. power: 96%

To continue with the simulation I need some more parameters:
From piston:
-Pin eccentricity.
-Compression height.
-Crown geometry.
-Weight (including pin, rings & retainers)
From connecting rod:
-Weight.
-Width (big & small end)
-Mid section shape “I”
From cylinder head:
-Valve angles (between valves, and relative to the cylinder deck).
-Combustion chamber “diameter” (shape al gasket level)

With the above parameters I will calculate the forces in the piston, pin, con rod, and bearings, that will be useful as a base to calculate and design the modifications of the components for the 3.8 version.

I wait for your collaboration (since my engine is still installed in the car and I cannot strip down it still)

Paco.

[Barryh]

If your going to run a turbo I would suggest not going much bigger than 3.4 L.
I would do this on a std. 72mm stroke 30l crank,not Gta 78mm.
3.7/3,8 l is marginal under boost as the heads dont seal well with that bore.
You have water ways very close to the sealing surface and the heads surface here warps,breaking the head gasket seal..This happens occasionally on the 3.8l N/A 156 engines,never mind turbos.

If you are going to build a big bore mator making 450-550 odd hp ,look at 72mm stroke and 98mm bore.

Another reason for staying away from 3.7 and 3.8 turbo is there will be nothing Alfa that you will NOT break with this engine.
Hence the front BMW gearbox rear diff we do.

Std crank will do for this engine.I would go for a set of priority rods here as the best GTA std rod will only be good to 7000rpm.plus I sleep better knowing I have good rods in there..

[FMG_V6_btb]

Thanks Barry for your recommendations about maximum reasonable piston diameter, that's something I was concerned about, more after seeing yesterday that pictures in the Alfa GTV6 forum, the piston diameter used in the engine was 100mm, and the liner sealing was too compromised in spite of the use of ribbed rims (I don’t know the exact name) as gas sealing gasket.

Also I was concerned about flame travel, piston stress, and knock tendency, because the big diameter. All together drives me to establish as maximum “reliable” piston diameter for that engine in 98 mm as you recommended.

About the crank… I’m not so confident with the original one, because I have found many V6 Alfa with rod-bearing-crank failures, I don’t know from where the problem comes, may be is the lubricating system, or crank vibration-bending under high speeds and loads, or maybe that problem happens to the rods (the load is not centred in the bearing due to the rod offset), or maybe the problem comes from the crankcase main bearing supports which haven’t the sufficient stiffness (that’s why I thought about the internal bedplate). In any case, I will study the bearing loads of this engine to have some objective values to make a better analysis.

In my case (application) I will not have any problem (I expect) with the Alfa gearbox or differential, because the engine will be installed in a proto car, which will use an Audi transaxle gearbox.

Paco.

[FMG_V6_btb]

The reason to design and build a new crank (apart of the technical ones) is basically the price of a new unit, here at Spain the price for a new 3.0 164 crank is around 2000€, and around 1800€ for a 156 GTA one. If you consider that build a new FORGED prototype crank by a German crank manufacturer costs around 3000€, I think is better the proto option. Also if you take into account that MAYBE more Alfa enthusiasts could be interested in that engine modification (as a whole engine or in parts), the production costs for the crank will be lower, maybe at the same price of the original ones.

I know that modifying the stroke to 82 mm implies design a new rod, but that is not so difficult or expensive.

I will continue posting results of the engine analysis or simulation, just only hope somebody find this interesting.

Paco.

[FMG_V6_btb]

Which kind of rod is better for my application? “I” beam?... “H” beam?... or maybe “+” beam?

The two driving parameters when designing a rod are stiffness and weight (let apart the titanium ones by its price), so… for the same weight which kind of beam is better?

I have seen in other threads 3 kinds of rods for Alfa’s V6, the original one which is “I” beam, the THS ones designed by Jim Kartalamakis, which is “H” beam, and the ones fitted by John van Rensburg (well, Dawie de Villiers) in his modified engines, which are forged from chrome molly steel and manufactured by Pauter Machine, (see picture below of the three kind of rods).

Well, I have made a simple (but not for so bad) analysis of the of the 3 different beam’s sections, assuming the same material for the 3 rods (see picture below). The three sections have the same surface, and consequently will produce the same weigh if manufactured with the same material. The “I” beam rod is selected as baseline because is the one fitted as standard in the Alfa engines.

The results show that the “H” beam is twice stiff than the “I” beam when tried to bend around the y axis, but is 44% less stiff when tried to bend around the x axis. The “+” beam is generally speaking the less stiff, 47% less around x axis, and 13% less around y axis. Of course all the above % can vary depending of the real geometry.

In fact the most important axis for stiffness is the x, because the rod tries to bend around this axis at high engine speeds due to the inertia forces produced by the angular acceleration and deceleration of the rod, the gas forces and the bulge.

For that reasons modern high performance engines uses “I” beam rods (picture: Ferrari V10 F1 rod).

Paco.

[slyalfa]

hmm from what I have seen I think posted here, is the alfa rods bend in the Y direction. That is rods that are permentley bent on inspection.

also if you take the H and incress the center some say untill it hits the x=100%mark where is the y?

[FMG_V6_btb]

To avoid misunderstandings about the bent axis etc. here you have a picture from a bent rod around the x axis (that’s the typical bending of the engine rods), so as I explained in my last post, “I” beam rods are conceptually stronger than the rest.

Of course, if you design a “H” beam rod with “long wings” the stiffness could be near of the “I” beam ones, but the rod width could result excessive and touch the lower part of the liner at high rod angles.

Paco.

[FMG_V6_btb]

Something about trapping efficiency (in this case can be interpreted as boost pressure) in-cylinder peak firing pressure, compression ratio, and power.

The following graph show different combinations of compression ratio and boost pressure (well, trapping efficiency) to obtain around 470HP.

The most important parameter here is the resulting in-cylinder peak firing pressure, because the engine durability depends of it.

In the 3.0 NA engine, the peak pressure (at max. power) was 92.6 bar, the durability limit of this engine could be around 110-120 bar, so compression ratios for turbo or supercharged engines higher than 9.5:1 could be disastrous for the engine durability. IMO 8.5 to 9:1 is the right value.

Paco