radiator efficiency

[tcslr]

is there any general rules of thumb to check the efficiency or performance of a radiator?

By running car until well and fully heated OR after some driving, measure the inlet and outlet temperature.  What delta is generally expected? 20F? 30F? 

[TexasJohn55]

After some driving, stop and raise hood, shoot temps, too late. It can only be done under load to be accurate. I believe 10* delta would be acceptable, less than that - marginal. 20-30* would indicate system more than adequate. DISCLAIMER: I have not checked my cars, only heavy duty diesels.   John

[TexasJohn55]

Once the temp reaches max thermostat open temp, you are done and lost control of system.  That would be 195-200* for a 180 stat. A good cooling system ( radiator, pump and air flow ) will consistently keep system at less than wide open stat temp.

[Rusty_OToole]

Temp gauge should tell you. If you don't have one you can install one. Start by blowing any dust and bugs out with compressed air from behind. See the fan is close enough to the rad or has a shroud. You can use an infrared thermometer to check the temp of the core, if there are cool areas it indicates blockage in those tubes. Look for corroded fins especially near the bottom of the rad. Look inside (when rad is cold) for lime buildup in the tubes or tank. What exactly is the problem?

[ojh]

I've heard that the 10deg delta another mentioned is the minimum acceptable as well, it will allow radiator to recover while on the road - the 10deg is stationary - you sit at traffic light and temps rise but recover when moving.

[Str8-8-Dave]

So delta T from radiator inlet to outlet is subject to numerous variables like effective radiator core airside fin and waterside tube construction, downflow or crossflow waterside routing,  core material, core area, airflow speed, air temperature, and airflow delta P across the airside of the core, water flow GPM.

 

Thermostats, either airside (variable shutter opening) or waterside are designed to control minimum water temperature, not maximum temperature.

 

Straight water boils at 212F at sea level unpressurized and the boiling point drops 3 degrees/1000ft elevation.  Differing coolants and varying concentrations may boil at higher or lower temperatures.

 

Pressurized systems IE systems run with a pressure cap boil about 3 degrees/PSI higher than unpressurized system, so a pure water system with a 9 psi pressure cap should not boil until it reaches 239 degrees (9 PSI X 3 degrees/PSI + 212).   

 

In radiator surface tunnel and vehicle testing we always used thermocouples in water inlet and outlet and air inlet and outlet of the radiator.  

 

I'm no subject matter expert but I did run radiator performance testing at Ford because we made them for Ford cars and trucks and sold them to other OEMs.  Most truck applications got copper/brass downflow radiators and cars got vacuum brazed aluminum radiators, some crossflow, some downflow for cars.  We had to prove radiators had sufficient capacity to meet the customers engineering specifications and keep their cool under all kinds of conditions.  These are just points to ponder.  In general our old cars use copper/brass downflow honeycomb cores and no pressure caps.  Radiators had to be oversized to keep them from boiling under load at high ambient temperatures.  If your cooling system can stay under 205 degrees worst case you are probably good.  In a pressurized system I would get nervous when top water gets to 10 degrees less than the calculated boiling point with a known pressure cap.  

Edited April 29, 2021 by Str8-8-Dave (see edit history)

[tcslr]

Thanks all.  My question pertains to unpressurized systems.  I realize that there are many variables that can have impacts.

I was running Chrysler and Chevrolet and measuring the inlet and outlet temperatures with a Fluke pyrometer.  These are both unshrouded fans. 50:50 antifreeze liquid.  The radiator cores are clean ( I beleive).

 

My hypothesis was that peridoically if I measured this and plotted over time, would I see a decay ( I'm sure) but likely over-heating would happen well before.  So what is a good REASONABLE metric? delta?

 

Allow me to give a similar 'rule of thumb.' In power generation one can calculate/estimate a plant heat rate ( typically very hard to do - many many variables) by using the turbine heat rate ( very well known) divided by the station service and boiler efficiency -- it gets one close. ( THR/((1-station service (%))*(blr eff (%)).   Simple!

 

Tom 

[Gary_Ash]

I looked into this a while back when someone was asking about a disk-on-tube radiator for a 1905 car.  Later engines have higher thermal efficiency, but 25 % is a decent estimate.  So, if you need 10 hp to the rear wheels at 60 mph, you need 40 hp equivalent heat in the engine or 40x746 watts = ~30 kW.  So, 10 hp or 7.5 kW goes to the rear wheels, about 11 kW out the exhaust pipe and 11 kW to the radiator.  If the radiator efficiency is in the neighborhood of 3000 W/m²-K - and that's a very round number - with an 18"x18" radiator core (~0.2 m²), then the temperature drop in the radiator is about 18 K or 32 °F.  So you could have 190 °F at the radiator inlet and 160 °F at the outlet.  At idle and lower road speeds, the temperature difference is less. On hills and during acceleration, the power load is higher, so the inlet water temperature goes up and so does the outlet.  You just want to keep it below boiling.

 

The 10 hp at 60 mph may seem low, but once a car is at speed, it doesn't take a whole lot to keep it there.  Older cars with the shape of a brick will need more hp at speed.  There isn't a lot of published data on actual car radiator efficiencies.

[Rusty_OToole]

Does anyone have any idea what this guy is driving at?

[Ken_P]

He’s trying to predict when radiator service is required, based on a decay in the temperature change across the radiator, before it is so bad/clogged as to cause the car to overheat. Neat question.

Edited May 1, 2021 by Ken_P (see edit history)

[TAKerry]

I certainly mean no disrespect to you engineers, and y'all are talking WAY over my head, but I keep an eye on the temp gauge, and check the fluid on a regular basis. All seems to work ok for me. 

[Str8-8-Dave]

5 minutes ago, TAKerry said:

I certainly mean no disrespect to you engineers, and y'all are talking WAY over my head, but I keep an eye on the temp gauge, and check the fluid on a regular basis. All seems to work ok for me. 

Now we're talking.  As a practical matter this is all that is required.  If it isn't boiling over under the worst of your driving conditions, if it isn't full of foam, losing coolant and if the temperature gauge isn't going nuts some engineer got his numbers right long before you got your car.  Drive and enjoy.

[ojh]

In a good many years of checking radiator temps the only radiator I ever saw getting more than 25degrees from inlet to outlet was on a 25 Nash, I'd be happy to see 15degrees on a driver, most are barely 10.

[ArticiferTom]

My '31 Dodge drops about 30 . There is a thread on servicing radiator with a Zoo Tool and CLR . Mine was more calcium build up . I show various IR photos there  Search ( Zoo tool ) . The tool was period service tool for rad systems .

[Larry Schramm]

I have always been told a good "rule of thumb" is 20deg difference from the inlet to outlet while at idle.  That number seems to work for me.  I have non pressurized systems.