Overheating BBC 468 Pro Street Willy's Coupe

[topradman]

I post this from time to time, but I'm not sure anyone wants to listen...
With cooling systems. People start guessing and start changing things. Some time you hit it, some times you don't. Guesswork gets expensive and time consuming. I can't guess any better than anyone else

There is some science to diagnose cooling system problems.

I couldn't agree with you more on this. I have been in the cooling/radiator business since 1981. This is number 2 out of 2 vehicles over the years that I am not figuring out what the problem is. The other was a blown 57 chevy, 540 Merlin w/Mooney Ham 12:71 and customer would not allow pulling the cylinder heads and he instead sold the car out of state.

Start with checking temps in at least four spots. I prefer to use thermistor probes into the coolant, an IR gun can be used to an extent.

IR guns are pretty useless on this car. You can't get a good reading through the grill ( 41 Willys w/oe style grille, chrome bars about 1/2" apart ), can't remove the grill with out draining system, moving wiring harnesses, removing the radiator/shroud/fan/overflow reservoir assembly as the lower radiator support has an integral air dam so cant get to the bolts ). IR guns are liars on shiny or even dull aluminum surfaces and even black hoses are considerably off on most IR guns. This car is PACKED under the hood so radiant heat from headers, engine block, etc interrupts good temp readings. If fan is on, it throws so much hot air that surface temp readings are not trustworthy either.

Place probes, take reading in at least four spots.
Radiator inlet and outlet. Before the thermostat and after the water pump.

Placing probes is not an easy endeavor either. Would have to remove the radiator, drill more holes and tig weld in more npt fittings, buy 3 more gauges to go with the one Autometer manual gauge ( $140 a pop for full sweep ) that I do have. ( don't trust electric gauges for serious diagnostics and don't trust "cheap" manual gauges as I have seen way too many variances in cheap gauges over the years. Gauge in the dash on this car is off 15 degrees or more past 190F and this is a new electric gauge that was in the car when it came to me ).

A good working cooling system will have a 12-15 degree delta T. ( across the radiator in and out, and also engine in and out )

Agree 100% with this.

Watch the reading as the engine is running and cooling system is working.

Temps Closer together than that indicate a cooling Capacity problem. (Insufficient cooling)

Temps Farther apart indicate a coolant Flow problem. (Restriction of coolant flow)

Closer = Capacity. Farther = Flow

Agree, however, if cooling system capacity is insufficient because of a mechanical problem such as the head gaskets being wrong for the cylinder head/block combo, the car might require a larger cooling package than can be installed which is really what I am beginning to question.

For example, If a capacity problem look at airflow and radiator size/capacity

If a flow problem look for plugged (internally)radiator, water pump issue, thermostat, collapsing hose, other flow issue

.

Shouldn't be a flow problem. Pump seems to move plenty of water, radiator is new ( I welded the tanks onto the core ) so I know there is no restriction there. Hoses are now 1 3/4" in & out including an upper water neck that is 1 3/4" in & out.


There is physically no more room available on the Willy's coupe without cutting frame horns off of a very pretty, powder coated gray metallic tube chassis and also cutting out inner fiberglass fender wells. Unfortunately, this is not my car or that would probably happen.

I feel I have a pretty firm grasp of what it takes to cool a properly running and properly designed engine and that's why I am here asking other now. Wow! I know, right?



If having trouble finding a flow problem, pressure gauges can help locate the restriction.

For the OP, ( and. Not criticizing) if you did this, you would be able to find out if you had to rework the radiator or airflow before you went ahead and did.

I appreciate your input and wisdom on this very much. If you or anyone else has any other thoughts I am all ears. I am going to roll a little more timing into the motor tomorrow morning and put a 140 amp alternator on to replace the 90 amp alternator, just for peace of mind, since the fan is pretty hungry on top of Meziere, HVAC blower motor, ac compressor and E-fuel pump just to take that out of the equation.

Again, any other thoughts and I am all ears, always!
Thanks,
Scott

[MadBill]

I post this from time to time, but I'm not sure anyone wants to listen...
With cooling systems. People start guessing and start changing things. Some time you hit it, some times you don't. Guesswork gets expensive and time consuming. I can't guess any better than anyone else

There is some science to diagnose cooling system problems.

I couldn't agree with you more on this. I have been in the cooling/radiator business since 1981. This is number 2 out of 2 vehicles over the years that I am not figuring out what the problem is. The other was a blown 57 chevy, 540 Merlin w/Mooney Ham 12:71 and customer would not allow pulling the cylinder heads and he instead sold the car out of state.

Start with checking temps in at least four spots. I prefer to use thermistor probes into the coolant, an IR gun can be used to an extent.

IR guns are pretty useless on this car. You can't get a good reading through the grill ( 41 Willys w/oe style grille, chrome bars about 1/2" apart ), can't remove the grill with out draining system, moving wiring harnesses, removing the radiator/shroud/fan/overflow reservoir assembly as the lower radiator support has an integral air dam so cant get to the bolts ). IR guns are liars on shiny or even dull aluminum surfaces and even black hoses are considerably off on most IR guns. This car is PACKED under the hood so radiant heat from headers, engine block, etc interrupts good temp readings. If fan is on, it throws so much hot air that surface temp readings are not trustworthy either.

Place probes, take reading in at least four spots.
Radiator inlet and outlet. Before the thermostat and after the water pump.

Placing probes is not an easy endeavor either. Would have to remove the radiator, drill more holes and tig weld in more npt fittings, buy 3 more gauges to go with the one Autometer manual gauge ( $140 a pop for full sweep ) that I do have. ( don't trust electric gauges for serious diagnostics and don't trust "cheap" manual gauges as I have seen way too many variances in cheap gauges over the years. Gauge in the dash on this car is off 15 degrees or more past 190F and this is a new electric gauge that was in the car when it came to me ).

A good working cooling system will have a 12-15 degree delta T. ( across the radiator in and out, and also engine in and out )

Agree 100% with this.

Watch the reading as the engine is running and cooling system is working.

Temps Closer together than that indicate a cooling Capacity problem. (Insufficient cooling)

Temps Farther apart indicate a coolant Flow problem. (Restriction of coolant flow)

Closer = Capacity. Farther = Flow

Agree, however, if cooling system capacity is insufficient because of a mechanical problem such as the head gaskets being wrong for the cylinder head/block combo, the car might require a larger cooling package than can be installed which is really what I am beginning to question.

For example, If a capacity problem look at airflow and radiator size/capacity

If a flow problem look for plugged (internally)radiator, water pump issue, thermostat, collapsing hose, other flow issue

.

Shouldn't be a flow problem. Pump seems to move plenty of water, radiator is new ( I welded the tanks onto the core ) so I know there is no restriction there. Hoses are now 1 3/4" in & out including an upper water neck that is 1 3/4" in & out.


There is physically no more room available on the Willy's coupe without cutting frame horns off of a very pretty, powder coated gray metallic tube chassis and also cutting out inner fiberglass fender wells. Unfortunately, this is not my car or that would probably happen.

I feel I have a pretty firm grasp of what it takes to cool a properly running and properly designed engine and that's why I am here asking other now. Wow! I know, right?



If having trouble finding a flow problem, pressure gauges can help locate the restriction.

For the OP, ( and. Not criticizing) if you did this, you would be able to find out if you had to rework the radiator or airflow before you went ahead and did.

I appreciate your input and wisdom on this very much. If you or anyone else has any other thoughts I am all ears. I am going to roll a little more timing into the motor tomorrow morning and put a 140 amp alternator on to replace the 90 amp alternator, just for peace of mind, since the fan is pretty hungry on top of Meziere, HVAC blower motor, ac compressor and E-fuel pump just to take that out of the equation.

Again, any other thoughts and I am all ears, always!
Thanks,
Scott

Thirty bucks will score you a four channel Type K thermocouple thermometer. You can wire up all four locations at once per Keith's excellent advice, which should get you to the bottom of the problem in short order: https://www.ebay.com/p/prime-Tc41-4-cha ... 1575527730
<UPDATE> It's under $25 here: http://www.ebay.com/itm/4-Channel-K-Typ ... 2210128131

[topradman]

Ok, thanks!!! I will order one now! So, do I not need 4, 1/8" female npt fittings in the system to install probes or are they surface probes? If surface probes, what do you think is the very best, accurate way to attach to a surface for an accurate reading? ( I did not take the time to look at the probes on your link, sorry ).

[stealth]

Nice find...I'm ordering one too....

[MadBill]

Thermocouple wire is very fine, say 20 Ga. or so. The sensing function is the temperature-sensitive micro voltage generated by the soldered junction of its dissimilar metals. You can either buy pre-fabbed Type K probes of the desired length or just get the connectors and a roll of the matching wire and silver-solder your own. Calibration is pretty easy: dip them all in boiling water and depending on your altitude they should quickly plateau at 209-212° F.

As far as attachment, probably easiest to just epoxy the junction tips to the rad end tanks in the required areas and peel them off after you're through. For the engine ones, again just epoxy them to the TS outlet and the exit region of the water pump.

The engine metal surface temperatures will be very close to that of the coolant, but to be really accurate you could cover the junctions with glass wool or header wrap 'bandages', which will ensure that the probe readings are unaffected by air temp.

An alternative which addresses the accuracy problem is to loosen the appropriate hose clamps and slide the wire in a 1/2" or so and re-tighten.

[topradman]

Just ordered the 4 channel temperature set up you recommended. I will post my results on here as soon as I have them. In the mean time, any other information from folks who have run into a similar problem and have solved it, I am all ears!

Thanks a bunch for the tip on the thermometer, that should help solve some similar problems in the future. I have temperature checking devices of several types but never looked for or knew there was such a thing as a multi channel set up like this where you could easily monitor in this fashion. I think I can use this like a single channel device and just run the thermistor inside the radiator hose and clamp it back down, then refill. Don't think it will hurt the leads squished between the rubber and the hose connections. I can at least get a simultaneous reading of inlet and outlet temps with out having to move the probes. That should be a very helpful diagnostic tool. It still won't tell me if the head gaskets are mismatched though.

[MadBill]

.. It still won't tell me if the head gaskets are mismatched though.

Just give me a minute to find the link for the illuminated 0.060" diameter fiber optic bore scope...

[topradman]

0.060" or 0.60" ?. Bill, your really are "Mad"!! ( LOL )... I really appreciate your time sir! I WILL solve this problem!!!

[Keith Morganstein]

Ok, thanks!!! I will order one now! So, do I not need 4, 1/8" female npt fittings in the system to install probes or are they surface probes? If surface probes, what do you think is the very best, accurate way to attach to a surface for an accurate reading? ( I did not take the time to look at the probes on your link, sorry ).

That 4ch looks good for the money! Easy to check accuracy in some boiling water too.

I would want the probes in the coolant. That unit in the link uses the wire type. You can sneak those wire type under the hose so it would work for the rad in and out that way. In the engine I usually use the metal probes with 1/8 npt. I'll drill and tap if there isn't a fitting to remove. I haven't tried the wire type into a casting. Maybe get creative with a barbed fitting, hose and clamp? (Or another thought). Drill a small hole through a 1/8 NPT plug and push the wire probe through it and secure it into the plug with RTV or epoxy. I saw extra probes on one of the eBay listings.

Back to the part about the head gaskets being wrong, making coolant flow the wrong path, or blocking it, etc... It's going to show as an issue, might need more probes as you diagnose, rear of head, front of head, etc...

Usually with the basic four probes you can fnd the area to look further.

Most of the cooling systems I troubleshoot are diesel. Fortunately those engine have lots of fittings. But still I'm making spots and sometimes on tough ones, I'll have 6-7 probes (and some pressure gauges) in before I'm done.

Im not too fond of IR guns either. If the surfaces are the same, as in color and finish, they can be useful enough to show delta T. I don't trust them for actual temp of the fluid inside.

[topradman]

Keith, that's a pretty good idea with the npt plugs and some type of high temp epoxy. Can't really get to the back of the heads but might be able to drill on the rear of the manifold to get to the water. I should be able to go into the middle of the heads (front to rear) through the factory temp sensor holes for a mid reading and just use the remote mounted autometer gauge for the front of the manifold collectively speaking. I ordered 2 of these 4 channel meters so I should be able to monitor up to 8 spots. I guess I have just been apprehensive to spend this much time in diagnosis and go to these extremes as I must be too used to "winning" with out this much hassle. I usually get it right the 1st time on a custom cooling system. MadBill has proposed a very good idea as have you. I will definitely post my results on here for everyone.

If it was my car, I wouldn't be as apprehensive. At some point, there will have to be an invoice made on this project. I donate time to customers ( especially street rod guys... ok, and also folks that are down on their dollars and need help as long as they appreciate it ) fairly regularly as I have resigned myself to probably working 'til I'm dead and I guess I am cool with that. As long as I am learning something new it's not so bad either. I have already "donated" about 30 hrs to this project now because the guy paid me good money for the custom radiator, radiator support, overflow, t-stat housing, etc., back in September of last year... and I feel very obligated to see this through.

I guess better to wear out than rust out, right?

[MadBill]

I should have mentioned that virtually all of the EGT dyno probes I've seen are type K (yellow plugs) with enough temperature range to handle most anything, so depending on how hard-wired they are, they perhaps could be borrowed for on-car testing. (Although smaller tips like 1/16" can place the junction closer to the heat source for better accuracy.)

Re the junction actually contacting the coolant vs. being a few sixteenths away through iron or aluminum*, trust me it make virtually no difference, especially if all are similarly mounted. If the coolant was at 200°, the outer surface of say the water pump inlet OD would be within 3-4°. A 1/2" wrap of glass wool overlay would reduce the delta to <1°. *Unless we're talking say head outer surface vs. a water jacket buried 1/2" or more deep.

[Old as Dirt]

Mr. Bill...... a wealth of information.. TY Sir..

[JoePorting]

Sounds like you need a bigger radiator.

[Keith Morganstein]

I see I duplicated some of what mad bill wrote.. Not intentionally and it's great to see a few of us thinking about these systems.

Once you start using the multi channel with probes, you'll never look back when diagnosing problems. A great investment for someone in the business. It seems like the long way at first, but it's not when considering the time spent changing things and re-testing.


I like that the multi-channel on eBay shows temps all together on the screen. The unit I use has a rotary switch and that can be a pain. (Need a pad of paper and pen to log it)

I have seen probes used on surfaces with heat sink compound. (A white paste) it's hard to keep them in place though.

One note, on capacity problems (size or airflow), delta T usually starts off pretty decent as it warms up. As the system gets hotter, temps get closer and closer together.

[fdicrasto]

I would love to see a picture of the engine compartment. Sounds like there is no room for any real airflow if that could be a factor. It sounds like a nightmare to work on. A pressure cooker on wheels. Just a thought. Hey how cool would hood louvres look on this ride.