Broken Alfa Nord Race Engines

Competition engines and ancillaries - general discussion
Guy Croft
Site Admin
Posts: 5039
Joined: June 18th, 2006, 9:31 am
Location: Bedford, UK
Contact:

Re: Broken Alfa Nord Race Engines

Post by Guy Croft »

Unlikely. The main jets are going to be sized around 145 = 1.45mm dia.

The reason why this or that cyl detonates first generally has more to do with exhaust manifold conditions and pressure state - this is highly variable and unpredictable.


G
rmerrell
Posts: 33
Joined: July 8th, 2011, 11:13 am

Re: Broken Alfa Nord Race Engines

Post by rmerrell »

Fair point, Guy - they're actually, as above, 165 (assuming that's 1.65mm).

Taking your point about variability & unpredictability, is there any consistency for a given manifold? I'm guessing not given that, despite the last 2 failures being on #2, there have been failures on #1 & #3 previously.

Coincidently (or probably not) my StSt manifold cracked at the flange weld on #2 as part of this latest failure.

I've bought the book by the way. Looking forward to reading it.
Guy Croft
Site Admin
Posts: 5039
Joined: June 18th, 2006, 9:31 am
Location: Bedford, UK
Contact:

Re: Broken Alfa Nord Race Engines

Post by Guy Croft »

There can be but I've looked online at pics of your ex header - it looks quite well thought-thru and I do not think it would per-se contribute to detonation.

The erratic nature of detonation failure (this cylinder failing or that one..) with a 'bad' header (ex manifold), can certainly be pinned on the header itself on occasions but there is more too it than simply that. Detonation is caused by the sudden autoignition of a region (sometimes called 'crevice region') of overheated gasoline/air mixture trapped in some part of the chamber - usually on the inlet side where the chamber is coolest. When it ignites it collides with the spark-generated flamefront leading to almost instantaneous excessively high pressure and massive release of heat. The charge is supposed to burn progressively with the flamefront expanding evenly (nicely?) across the chamber. Detonation is just like an explosion, not progressive at all and that is why it punches thru the nearest, weakest part of the engine. That massive pressure has to go somewhere.

Engines like yours and many other 8V units all have pockets/regions of charge lurking but they don't always detonate. More often than not the unburned charge does no real harm other than put up the HC emissions by simply going out with the exhaust gas. All engines have piston-induced swirl - vortices created as the rings scrape up the bore on the compressions stroke, but engines like the 8V Fiat TC which have squish bands to compress the charge between them and a flat piston crown have squish-induced swirl too and they exhibit better charge homeogeneity (mixing of air and gasoline) and burn more efficiently than your engine (which has no squish bands at all). And you don't have any port-induced swirl either. The chances of a build up of fuel in 'corner' of the chamber (for want of a better word..) which is liable to 'flash off' increase the fewer of those combustion-enhancing features you have.

Add in the intrusion of a piston dome right in the middle of the flamefront and you can see another reason why the flamefront could well fail to ignite the charge effectively and the bigger the intruder dome the greater the risk. An overly large intruder not only slows down the flamefront (ie: the burn) it naturally increases the risk of detonation. It will be obvious to those reading this that there is a trade-off between burn and CR and you may well have gone too far in favor of the latter.

So some reasons why unburned charge exists in the chamber at all, but why/when does it detonate? It can certainly happen if the fuel octane is insufficiently high for the CR. But determining the RON rating required is a minefield in itself. Let's suppose we know a guy who ran an engine similar to yours (but not quite identical in every way..) with no detonation at all and we used the same stuff and got detonation. Why have we got it and he hasn't?
(edited by GC)

On an atmo engine the routine compression pressure even with 13/1 CR is nowhere near high enough on its own to cause autoignition of the crevice regions (this isn't a diesel) - the problem is overheating of the unburned charge region. Leaving aside the header the primary culprits - in my broad exp - are as follows:

1. Intake air too hot - can cause random detonation due to small variations port-port. It's vital to induct from atmosphere via a sealed, ducted intake path with front mounted scoop.
2. Ignition timing not optimised. I cannot say over-advanced or retarded because they are both equally damaging. The ONLY way to determine the optimum is dyno test with distributor swing at max advance. But for the test to be meaningful the intake air temp has to be optimal (above) etc etc etc etc.
3. Spark plugs too hot - I cannot assert that your plugs are totally wrong but with the kind of intended power you seek you should probably be on NGK B9EGV which are a gold-palladium fine wire race plug. I have been an NGK race distributor for many years and was a race distributor for Champion for many years too and plug selection is something I take very seriously. An overheating plug can create a hot-spot which causes pre-ignition of the main charge before the spark itself. This pre-ignition phenomenon can create local and sustained overheating in parts of the chamber. However it is generally rare to see terminal detonation from this fundamental cause and a simple examination of the plugs after a full load test will tell you if they are giving problems or not.
4. Excessive system back-pressure due to restrictions in the silencers. Race types MUST be big-bore straight thru. Exhaust back-pressure prevents the cylinder evacuating properly and you can get weird reverse-flow effects between cylinders. It can be very degrading, damaging ex seats and valves, upsetting the jetting (everyone jets up way rich to try and compensate - you can't...) leading to bore washing. Very similar in effects to 5) below.
5. Repeated exposure of one part of the chamber to excessive temperature. I'm not talking simply about intake air here. This is not often discussed. Here I am referring to what can happen if the exhaust event is ineffective. At TDC on the overlap phase (discussed in my book as you will read) the idea is that incoming charge is physically encouraged to flow over the top of the piston and out of the exhaust port - cooling everthing in its path and at the same time purging the cylinder of residuals - leftovers from the previous (ex) phase. This is an absolutely VITAL part of how tuned atmo engines work and if it is ineffective it is both 'power-sapping' and dangerous. Why has no-one ever succeeded in getting over a genuine dyno-certified power figure of 260 on the Sierra Cosworth head? Because the ex ports are tiny and restrictive even with best practice at enlargement and the best E/I you'll ever get is about 65%. I've done it. It's just not enough! I've proven this ratio over and over. Many heads including virtually every cast-iron head ever produced in the UK for production sports cars since WW2 are the same. Some designers just never understood this. If the overlap event is ineffective you are certainly going to get an ongoing heat problem in the chamber somewhere. That event is partly a function of ex valve size and partly ex valve lift but importantly does the port region itself flow enough? It's what, only 2" long or so and could it really upset everything? You bet it can.

Having a big ex valve and a high lift cam and a good header are important things but it is not everything. The header can only flow what's fed to it by the ex valve. And the particular head you have exhibits - in standard trim - very poor flow in the ex port. Whether all the Alfa experts know this I couldn't possibly say but I certainly do. There is a strict minumum ratio of ex-inlet flow you must try to attain and that is 70% E/I bare port flow (ie: measured without valve) and the standard head of your type is nowhere near this. The valve-valve area ratios are right but frankly the as-designed ex ports are a complete disaster and someone at Alfa slipped up badly in that area. Modify the inlet ports for more flow and you make the problem worse. You must start the head mods with the ex port and get that flow up or the cylinder is never going to purge properly. And can all this lead to random detonation? Indeed it can. Though asserting saying how random it really is - or whether you just 'caught it' in time (as the saying goes) - is another matter entirely; most engines with extreme detonation either stop or get switched off pretty quick. If they were able to carry on running you might well find all the cylinders did it.

I said above, "Let's suppose we know a guy who ran an engine similar to yours (but not quite identical in every way..)" and there what I was referring to was maybe he knew what I do about the ex ports on your unit.

No doubt other well-briefed readers can add to this, it's very hard to be exhaustive (pardon the pun) in the time I have but I hope this helps some.


G
Attachments
check out the before-after exhaust BPF and that I reasoned (quite rightly) that for reasons of E/I and power target there was NO point at all 'aiming for the skies' on the inlet flow. All things being equal even 112cfm could take a good 8V 2 liter engine to 200bhp.
check out the before-after exhaust BPF and that I reasoned (quite rightly) that for reasons of E/I and power target there was NO point at all 'aiming for the skies' on the inlet flow. All things being equal even 112cfm could take a good 8V 2 liter engine to 200bhp.
2 liter 1973 Alfa 8v TC flowtest.JPG (78.38 KiB) Viewed 12400 times
see what I mean. Look at the guide loss!!!!!!!!
see what I mean. Look at the guide loss!!!!!!!!
Alfa flowtest std ex port with guide in a dismal 65.4cfm @ 10in.JPG (528.96 KiB) Viewed 12400 times
Alfa flowtest std inlet port - even with guides in a whopping 104cfm. That's an EI ratio ofonly 63% which is hopeless.JPG
Alfa flowtest std inlet port - even with guides in a whopping 104cfm. That's an EI ratio ofonly 63% which is hopeless.JPG (114.33 KiB) Viewed 12400 times
if you think the Alfa TC is bad check this one out. It's not exactly 'rocket science' to get the final E/I on the Alfa up to 83%
if you think the Alfa TC is bad check this one out. It's not exactly 'rocket science' to get the final E/I on the Alfa up to 83%
Sierra Cosworth full spec flows.JPG (102.16 KiB) Viewed 12399 times
rmerrell
Posts: 33
Joined: July 8th, 2011, 11:13 am

Re: Broken Alfa Nord Race Engines

Post by rmerrell »

Wow - thanks, Guy - that's given me some food for thought (and made me even more keen to ambush the postie to get my book).

Starting with thoughts as they arrived in my head from reading what you've written: Is there some guidance in the book regarding cold air feeds? I think I have a reasonable understanding of the basics of these systems from other reading that I've done:
- carefully designed inlet
- large (100mm ID?) smooth inlet tract
- large plenum chamber to recover pressure from fast-flowing air

There are a few challenges not least of which is the lack of space between the inlet and the inner wing & steering box in my car. That said, the most "experienced" 105 racing specialists I know told me that they have failed to build an air feed system that does anything but sap power. I'm keen to try to do better.

The other key question that I have is in relation to exhaust flow. I think I've understood (will read more to confirm) what you've written about the Alfa head and exhaust flow. That said, what I've both read & seen in practice on seemingly successful Alfa heads, is tuners fitting smaller exhaust valves (whilst still porting the exhaust tract). Why would that be? Was it simply already way oversized for exhaust tract (even in it's enlarged form)? Again - if this topic (obviously not as related to Nords but in general) is covered in the book, I'm more than happy to wait and read.

Thanks again, Guy - you've successfully got my brain whirring on all of this stuff (I wish I'd been better at fluid and thermodynamics on my degree!). As much as I enjoy steering the car around the track, I love the preparation and improvement. What racing has done for me is given me the drive to improve and therefore the drive to learn.

Thank you!
Guy Croft
Site Admin
Posts: 5039
Joined: June 18th, 2006, 9:31 am
Location: Bedford, UK
Contact:

Re: Broken Alfa Nord Race Engines

Post by Guy Croft »

Thanks for your kind comments.

There is an optimum ratio of ex-inlet valve area same as the relative port bare flow and that is as above, 70% and as high as 80%, not more. There is no good argument at all for making an ex valve smaller if it meets that criterion. If you look at almost any engine designed in the last 20 years 8 or 16v you will see this is now quite widely known. The reason an engine will work of course with an ex valve smaller than inlet is the higher pressure ratio across the valves at initial opening and in the later phase of the ex cycle the pumping-out work done by the piston.

G
rmerrell
Posts: 33
Joined: July 8th, 2011, 11:13 am

Re: Broken Alfa Nord Race Engines

Post by rmerrell »

Thanks, Guy.

I'll see see if I can find the rationale stated elsewhere for reducing the Nord engine ex-valve (at the same time as increasing the in-valve) and share on here to see if it stands up.
Guy Croft
Site Admin
Posts: 5039
Joined: June 18th, 2006, 9:31 am
Location: Bedford, UK
Contact:

Re: Broken Alfa Nord Race Engines

Post by Guy Croft »

OK but it won't interest me any!

Look hard on forums where other spurious mods like 7mm valve stems, 46mm inlets, titanium valves and right-foot dyno bhp are discussed.

G
WhizzMan
Posts: 459
Joined: August 13th, 2010, 8:05 pm
Location: Amsterdam, the Netherlands

Re: Broken Alfa Nord Race Engines

Post by WhizzMan »

I can confirm the in-ex ratio. Alfa has had a tendency to oversize it's exhaust valves on production cars. There must be a torque band, emissions reason or something for that, because their racing departments know perfectly well to mostly increase the inlet valves in size on the racing versions of those engines.

Guy, could it be that decreasing overlap and compensating with larger valves will keep the torque band wider and lower in the rev range?
Book #348
Guy Croft
Site Admin
Posts: 5039
Joined: June 18th, 2006, 9:31 am
Location: Bedford, UK
Contact:

Re: Broken Alfa Nord Race Engines

Post by Guy Croft »

There is no substance to a debate on valve size per-se on this type engine in my expert opinion, study the data below from one I did recently. Look how the critical E/I ratio can be balanced-out with judicial work on ports and seats. Going to a smaller ex valve would necessarily mean a smaller throat and you just simply could not achieve anything like my end result by that means.

The ex valve size in itself (ie: per-se) means little. One must examine the port/valve behaviour to determine the effectiveness of both as a system/pairto make a judgement re their impact on the exhaust event.

Moreover restricting an ex valve size simply on the grounds that a % of inlet area it 'seems' to big is likely - again - in itself - likely to inhibit the cross-flow valve to valve in the very overlap phase that we must augment to gain top-end power. Indeed if we don't work on that the top-end power and spread of torque is going to be useless, partly for reasons mentioned in my earlier post (purging/cooling - charge density).


G



HEAD DATA BEFORE (mm) HORIZONTAL VERTICAL
IN PORT OUTER SHAPE CIRC
SIZE AT FACE 38.3
IN PORT OUTER AREA APPR sq mm 60.2
IN PORT SMALLEST SECTION 34.4
EX PORT OUTER SHAPE CIRC
SIZE AT FACE 32.0
EX PORT SMALLEST SECTION 28.6
VALVE IN DIA/STEM 44.1 9.0
VALVE STYLE INLET OE 30 x 4.55 x 18 deg x R21
IN VALVE SEAT 30 x 3 x para
THROAT IN 39.1
VALVE ANGLE
VALVE EX 40.0 9.0
EX VALVE SEAT 30 X 3 X PARA
THROAT EX 33
VALVE ANGLE
GUIDES OE
EX-INLET THROAT AREA RATIO 71.2%
EX-INLET VALVE AREA RATIO 82.3%
E/I FLOW STD 62.9%


HEAD DATA FINAL (mm)
IN PORT OUTER SHAPE CIRC
SIZE AT FACE 39.7
IN PORT OUTER AREA APPR sq mm 62.4
IN PORT SMALLEST SECTION 35.0
EX PORT OUTER SHAPE CIRC
SIZE AT FACE 33.0
EX PORT SMALLEST SECTION 31.0
VALVE IN DIA/STEM 44.0 8 STEM
VALVE STYLE INLET 45 x 3 x 18 deg x R21
IN VALVE SEAT 20 x 45 x 2mm x 70 blended around SSR
VALVE EX 40.0 9 STEM
EX VALVE SEAT 15 x 45 X 2mm X LIGHT 70 INTO PARA
THROAT EX AS STD
GUIDES bronze 14mm nom OD client supplied shortened by exhaust 12mm off nose and inlet 5mm off nose
EX-INLET VALVE AREA RATIO 82.6%
E/I final spec port with guides 72.5%
rmerrell
Posts: 33
Joined: July 8th, 2011, 11:13 am

Re: Broken Alfa Nord Race Engines

Post by rmerrell »

Guy - are the 8mm valve stems you used on this engine for the in-valve for flow purposes or valve weight?

The book arrived yesterday BTW - it's great - my brain was full last night! Got me thinking about X1/9s too (my first proper car was a 1500 - I've also had a Fiat 20VT) as I've always wondered what sort of race car they would make in our series.
Guy Croft
Site Admin
Posts: 5039
Joined: June 18th, 2006, 9:31 am
Location: Bedford, UK
Contact:

Re: Broken Alfa Nord Race Engines

Post by Guy Croft »

Funny you should ask - the valves just happened to be that way - the client kindly supplied them!

'Always work with what you have in front of you' *

GC



* Rule 9 from the GC 'Book of Life'
rmerrell
Posts: 33
Joined: July 8th, 2011, 11:13 am

Re: Broken Alfa Nord Race Engines

Post by rmerrell »

That makes sense! I understand that their key value is in the weight reduction (I saw in your book that the V-Tech uses 5.5mm!) as I can't imagine a mm thinner stem has much influence on flow in that particular location.

I have most of my last engine available for re-use so I'm keen to take my time and apply the principles I'm learning here to that motor. Much of the "hard" work is done - I think it's a fundamentally sound engine which has been ported on a flow bench. The places that I'm learning it was flawed are:

- CR far too high. I believe I can fix that with some less ambitious pistons (it was running 12.5:1 JE "big valve" pistons with large intruders)
- Timing too advanced
- Venturis too small

It had previously reasonably reliably delivered 173bhp @ 7200rpm. It needs a full strip down and clean and the crank needs a polish. But, even if it's just a spare engine, it will be reasonably inexpensive.
rmerrell
Posts: 33
Joined: July 8th, 2011, 11:13 am

Re: Broken Alfa Nord Race Engines

Post by rmerrell »

Guy - can you help me understand the way in which having a shorter duration exhaust cam than inlet (320/300 in my case) helps deliver torque? I've always questioned why inlet & exhaust should be the same shape & duration (after all they are dealing with mixtures/gases that are quite different). I started with cams that were both 320 but the tuner discovered on his dyno that a shorter exhaust duration delivered more torque for no significant loss in power.
timinator
Posts: 116
Joined: March 9th, 2011, 5:20 pm

Re: Broken Alfa Nord Race Engines

Post by timinator »

[quote="rmerrell"]Guy - I started with cams that were both 320 but the tuner discovered on his dyno that a shorter exhaust duration delivered more torque for no significant loss in power.[/quot


Horsepower is calculated by a formula using the engine torque output. If you have higher torque it follows that horsepower must increase. 'No significant loss in power' is not a possibility. As you have already witnessed it is easy to install a cam that does not work well in your engine when you have no experience to base that selection on. Your experience though is increasing. Asking for an explanation of cam timing is really to broad a question for you right now since you didn't recognize that GC was answering that question in his reply to you. Do some reading on the subject and you will find out what I mean.
Could you supply the air/fuel ratio graph for the dyno test you posted as it would help explain more of what your engine is experiencing. A graph of engine timing would also be helpful.
A picture of your domed piston would be useful too. If you are are not familiar with BB Chevy engines I can assure you they have massive domed pistons for their 119cc combustion chambers. It is common for these engines to run 12.5:1cr with 40+deg. of timing.

Good luck with your project.

Tim
WhizzMan
Posts: 459
Joined: August 13th, 2010, 8:05 pm
Location: Amsterdam, the Netherlands

Re: Broken Alfa Nord Race Engines

Post by WhizzMan »

A lot of times when people claim they "get more torque without losing horsepower", they are in fact telling that they get more horsepower/torque in a wider rev range, without losing the absolute top horsepower figure. Could this be the case here as well?
Book #348
Post Reply

Who is online

Users browsing this forum: No registered users and 43 guests