project 1 from Ogorir: ducting in a radiator

one of the cars I'm currently working on is a '63 Saab 96. originally, these had the Saab 3 cylinder two-stroke with a tiny little radiator mounted above and behind the engine and a rear-facing shaft-mounted belt-driven fan. they were notorious for overheating, particularly with ethanol blended fuel (the mixture runs leaner the more alcy is in the fuel, and hence hotter).

this car, though, as the V4 engine Saab used in the later 95/96s and sonnets. it's a few inches longer than the 2-stroke, so it's really jammed in the engine compartment. I didn't do the engine swap on this car; it was left half-done, which is how the boss bought it. he had a the local VW beetle shop get it running again, but it still had the stock radiator in it.

the boss bought a used radiator from an '86 Saab 9000 on ebay a while ago, which I'm currently making bracketry and ducting to mount it in this 96.

I've deleted the heater core and fan blower unit off the firewall so I have more room to work with. they're not really necessary here for 11 months out of the year.

the 9000 radiator is 16.5"h x 27"w, so I had to lay the radiator down to about 25 degrees off horizontal to fit the top hose under the hood. because of the angle and it's position behind the engine, I'm ducting the exhaust side out the wheel wells so it doesn't dump back out and heat the radiator back up. there are two engine air outlets ontop of the inner fenders that I've taken out to allow more air to pass. the center section between the shock towers needs to be removable for access to the shift linkage and steering rack.

I've made the tops of the duct boxes removable (well, one. I still have to make the other one). I'm going to have to run wires, lines, ect through the duct area as the duct plenum occupies pretty much the entire area behind the engine from fender edge to fender edge.

eventually, this car will get a megasquirt 3 ECU, controlling port EFI and electronic coil-per-plug ignition. charly wants to stick a turbo on it, too. if that's the case, I'm going to have to re-do all this stuff, as the only place to stick the turbo is right behind the engine, which would heat-soak the radiator for sure.


It's a fun project, though. I'm currently doing another radiator swap in one of the long-nose 2-strokes, as well as changing to an electric water pump. they have a typical front-mounted radiator, but they were still only 12x12, so not a whole lot of cooling going on there. the core on the new one is 21wx14h and will have a 14" electric fan on it. there's not going to be near as much fabrication on that one, else I'd do a project out of it, too. it's a higher priority, as it's a complete car, so the V4 car is sortof on hold at the moment. this car is also getting a new engine, mostly 'cause charly just got a hopped up one back, not because it really needs a new engine. it really just needs a new harmonic balancer.



Pictures:
This is the stock engine compartment (this is actually a different car, but the V4 looked the same before it was v4'd
Attachment 3605
these two are from another short-nose that the boss bought recently. body's in good shape, but there's a good deal of bondo in the bolt on panels and the mechanicals are mostly toast. we have plenty of good doors and fenders and a few rebuilt engines and trannies coming back soon, so no worries!
Attachment 3607Attachment 3609
left and right view w/o radiator. the silver everywhere in these shots is just silver pain, not bare metal. it was starting to flash and I was out of acid etch. the rebar 'stud' in the top center fits through a rubber grommet in the radiator. originally it would've been a bolt, but one of the mounting studs on the bottom of the radiator is broken, so I put the bolt in that corner.
Attachment 3604Attachment 3601
with radiator in place.
Attachment 3602Attachment 3610Attachment 3603Attachment 3608
you can see the alternator pulley cut into the hood and the timing cover peeking out in this shot. it's a real tight fit up front.
Attachment 3606

I thought I had more closeups. I'll take some today. as I said, this project has been put on hold to finish the long nose engine swap. I put the new engine in it yesterday and started fitting the big new radiator in that one yesterday. biggest hurdle left on that car is figuring out the best water flow with the new electric water pump. the stock pumps inlet is off the front of the head and it has 2 outlets, one feeds the bottom of the radiator, the other feeds the block water jackets.

Does you boss resale the Saab's that you restore? How much horsepower does a 3cyl 2 stroke put out?

No. he's given one to one of his son in law's, though. there are 8 that drive, including the long nose that's getting the new engine.

stock, they made 44hp. fully tuned stage 3 they make 94-98. there's a guy in sweden who makes stroker crank webs, so we might be getting some of them in the future. fully bored and stroked, they punch out aroun 1100cc (up from 800). I'd really like to run electronic ignition and fuel injection on these, though. I think we can pick up a few easy HP and keep them from killing themselves when you let off the throttle at high speed (mixture leans out, less oil on the cylinders, temps spike. it tends to blow a hole in the center piston)

the V4 in stock tune made 65hp, mild tune is 100, max normally aspirated is around 140. Saab's best turbo effort was 240hp, but it was deemed to unreliable for rally use. they detuned it to 200hp w/ the turbo. I'm aiming around 225, if we go the turbo route. engine control and turbo theory has come a LONG way since the early 70s.

What kind of fuel system is on these engines that they can't be tuned not to run lean? Sounds like poor tuning too me.

Do you have any pictures of the project?

triple single barrel solexes. the issue with them running lean is at high speeds and you take your foot off the gas partway, the engine is still spinning fast, but you've reduced the flow of air and fuel(and oil), so the combination of all three causes the combustion temps to spike. these cars have a freewheel, so once you let off far enough, the transmission disengages and the engine idles.

so, to compensate for that, atleast in some part, you have to run them really rich. not the end of the world, but even run at 11.5:1 air/fuel, you can still lean it out far enough to hole a piston if you drive it wrong.

yes, I do have pics. I forgot to go out to the office and post them. I guess I'll go do that.

Thanks for explaining, I can see where the freewheeling would contribute to lean problem. Most of my experince with two strokes is motorcycles and wave runners not cars I lean towards 60 an 70 american muscle car which are much easier to tune. Thanks again for trying to teach an old dog .

it's actually the freewheeling that saves them. when the freewheel kicks in, the engine idles, so it's not spinning at cruising speed (4500rpms, give or take), so the oil supply is adequate. it's that spot between when you start letting off the gas and before the freewheel lets the engine RMP drop that's the issue. mostly, you just can't ease off the gas while cruising. you have to get OFF the gas and let it coast. I'm hoping with FI I can rich it out when you close the throttle at high rpms, thus keeping the oil supply adequate. from my understanding, it's a combination of lean running (due to higher scavenging at high RPMS) and increased piston skirt friction that increases the combustion temps.

I've never tuned a megasquirt, so I don't know if I can be that specific. I don't want it to rich out on the way up :)

these 2-strokes are definitely an odd beast. their saving grace, in my opinion, is the fact that they weight about 120lbs fully loaded. I only use the engine crane about 1/2 the time to throw a new block in. lining up the main shaft in the clutch splines and the pilot is a bear, so I end up picking the damn things up manually even if I drop it in with the cherry picker.

Wow I can safely say I have never seen anything quite like this, I can see way they run hot. If I understand it depends for air from small scoop at the base of the windshield which could be problem as there is a high preassure area at the base of windshield, thats why 70 ss chevelle had cowel induction scoop faced windshield. Just a thought.

there'll be a big pusher fan on top of that radiator, so the 'natural' flow of air won't matter a whole lot. the main and side grilles in the hood and the slot grilles towards the rear of the hood will be the main air inlets. the airflow through the radiator is ducted out the tops of the wheel wells.

actually, that scoop you're looking at in the cowl is the cabin air intake. speaking of, if yuo look at the non-radiator pics, theres a big hole in the firewall. that's where the fan/heater blower unit bolts. on the inside, the air routing box bolts to the same studs. I have half a mind to put the inside box back on so you still have heat and defrost, but it would be powered by the radiator fan instead of a separate blower!

and, to be honest, there IS nothing like this. the engine compartments on these cars are TINY. the space between the wheel wells is 30", shock towers are about 27" 1/2way down.
interestingly enough, the only metric stuff on the V4 cars is the ford engine and the bosch alternators.


Edit: I forgot to mention, the V4 that's in these is a ford cologne engine, the V6 version powered the ford ranger through '91. original application was the ford taunus.

You should restore a rally version or build one, I bet that would be sharp!

possibly in the works, Sean. I have to get these couple of hopped up street drivers done first, if the boss ever stops breaking the daily drivers. daily.

here's the other car I'm sticking a big radiator in:
Attachment 3655
the big radiator:
Attachment 3652
bottom mounting studs:
Attachment 3647
starting the fan ducting:
Attachment 3649Attachment 3648Attachment 3650
tack welding the splits for a bit more strength(note how much it warped after welding). even just tacking those cuts brought the strength back almost to where the panel was before the holes were cut:
Attachment 3651Attachment 3654
the radiator in the hole:
Attachment 3646
I modifed the back plate from a water pump to make a second water neck to feed the remote electric pump:
Attachment 3653

(more pics next post)

water neck again:
Attachment 3658
the bracket that holds the top of the radiator to the grille. I'm going to need to weld another bracket to that one to triangulate to the inner fender. originally there were 2 angle steel brackets that did the same thing, but the corners of the radiator are in the way now, so I need to fab new ones:
Attachment 3660
you can see that the water pump has two outlets. one feeds the bottom of the radiator, the other feeds the water inlet at the bottom of the block.
Attachment 3657
I made this y connector to split the output from the water pump between the two inlets. it might've been moot, as I may end up running everything in dash 16 AN hose and using an -AN T fitting, but I have to run that by the boss. the fittings and hose to do that will be close to $350.

I've started the search for molded hoses I can re purpose, I think I've found a few, but they're all $30, and we need 4, so even standard hoses are looking like $120+
Attachment 3659

this is a stock engine compartment:
Attachment 3661Attachment 3656

a few pics of the other car of the radiator attachment point details. you can see the broken stud corner, the stud corner (wrapped in electrical tape for a temporary grommet), and the rebar pin up top. I might replace that with a bolt at some point.
Attachment 3664Attachment 3662Attachment 3666

also a better shot of the drivers side inner fender where I opened up the holes for better outflow. this is under the cover plate on the drivers side in the other photos. I still need to add a brace back across the hole:
Attachment 3670
an idea of the parts yet to be reconditioned:
Attachment 3663Attachment 3669
Attachment 3665Attachment 3671
and a pile of newly cleaned carbs awaiting new gaskets and re-assembly:
Attachment 3668Attachment 3667