Swap Science: Keeping Things Cool by Building a Cooling System

By Carl Heideman
Nov 27, 2018 | Mazda, MG | Posted in Shop Work , Drivetrain , Features | From the May 2017 issue | Never miss an article

Photography by Carl Heideman

In TV shows and web series, engine swaps happen in no time flat. Hammer this, weld that, and then launch the car into the sunset with a big, smoky burnout. High fives follow. Fade to black. 

These fast-paced transplants can make for good entertainment, but they typically don’t last for the long haul. We prefer to engineer swaps that can endure years of enthusiastic use, and that’s what we’ve been doing with our Miata-powered MGB project. We spent time designing engine and gearbox mounts that properly cradle the engine. The clutch now has a smooth linkage. The throttle cable should provide years of faithful service. We even thought out the charging system well enough to leave room for the occasional alternator replacement.

Now it’s time to consider the cooling system. Just slap a radiator in place and let the credits roll, right? Well, we’d rather have a setup that won’t leave us stranded on the side of the road. We also want parts that we can source easily, both now and in the future, and that means sticking with Miata and MG parts.

Steps 1, 2, & 3:


We’ve installed our 1.8-liter Miata engine in the body of a 1967 MGB, and to keep it cool we’re going to use
a little local knowledge: 1977-’80 MGBs came from the factory with their radiators mounted a little farther forward in the chassis. We’re going to duplicate that cooling system setup to provide a little more room for our transplanted engine. 

While the radiator mounts for these 1977-’80 cars are available from suppliers like Moss and Victoria British, we chose to make our own using a technique called hammerforming. (We detailed this process back in our November 2003 issue; visit the Features section of grassrootsmotorsports.com to read the full story.) Note that we haven’t drilled holes in our hammer-formed mounts. That step, we’ve found, is best left for after installation.


While the radiator mounts for these 1977–’80 cars are available from suppliers like Moss and Victoria British, we chose to make our own using a technique called hammerforming. (We detailed this process back in our November 2003 issue; visit the Features section of grassrootsmotorsports.com to read the full story.)


Note that we haven’t drilled holes in our hammer-formed mounts. That step, we’ve found, is best left for after installation.

Steps 4 & 5:


After fabricating our new mounts, we used measurements from another MGB to decide where to clamp them in place. We aligned the clamps with each other using some angle iron; an angle finder ensured that the mounts were perpendicular to their base.


Step 6:


We marked the locations of the mounts so we could remove them and drill the holes needed for the plug welds that will secure them.

Step 7:


Once we welded the mounts in place, we clamped the radiator in position. Then we marked the radiator’s mounting hole locations, unclamped and removed the radiator, drilled the holes, and bolted the radiator, back in.

Steps 8 & 9:


Time to address the fan, and this is where sticking with our factory parts strategy paid off: We simply ordered a complete set of fan components for a 1977–’80 MGB and bolted them in.


Step 10:


Next we had to hook up the hoses. The Miata’s stock lower hose setup features a pipe in the middle, and a previous swap taught us that we could retain it with only a minor adjustment to the length of the factory hoses.

Step 11:


It’s hard to get a picture of the whole installation, but this shows how the hose and pipe work their way to the lower radiator outlet.

Step 12:


For previous swaps, we’ve made a little bracket to secure the pipe’s mount to the car itself.

Step 13:


However, this time we had some interference. Thank the air-conditioner compressor bracket that was in the way.

Step 14:


We decided that a band clamp with a spacer could do the job. A 10-32 screw would hold it in place.

Steps 15 & 16:


We used a nut insert tool to put a 10–32 bung into the MGB’s frame rail.


Step 17:


The lower hose installation worked out well, and the bracket will do a good job of isolating engine shake from the radiator. By the way, shaky radiators are an issue with a lot of swaps, and at-risk radiator components usually reveal themselves after a few hundred miles with the new engine.

Step 18:


The upper hose setup looked pretty straightforward, as it just needed to cover a pretty short distance. The problem was our strategy of using only MGB or Miata parts: Initially, none of the hoses we looked at could make the link.

Step 19:


Success! We finally realized that a section of a lower hose for a 1968–’76 MGB would do the job. It just barely fit our strategy, but it worked well in the end.

Step 20:


Finally we had to deal with filling the system. We got a bung and cap from Summit Racing and had a radiator shop put them on our MGB radiator.

Step 21:


Our last step: installing the stock MGB expansion tank, which also houses the system’s pressure cap. We put the tank on the driver’s side fenderwell.

Read The Whole Swap Science Series:

  1. Swap Science: 14 Steps Before You Swap That Engine
  2. Swap Science: How To Build Custom Engine Mounts From Scratch
  3. Swap Science: How To Make Transmission Mounts
  4. Swap Science: How To Get Your Clutch Setup Just Right
  5. Swap Science: How To Make A Functional Gas Pedal
  6. Swap Science: How To Hang An Alternator
  7. Swap Science: Engine, Steering, and Exhaust Fitment
  8. Swap Science: Keeping Things Cool by Building a Cooling System
Join Free Join our community to easily find more Mazda and MG articles.
11/13/18 2:32 p.m.

I've gotta say I'm impressed with these bite-sized posts dealing with engine swap issues. They're excellent ways to break down the process step-by-step, and provide detailed information on what to consider doing, and why.

I'm curious here, though, about what you're going to do about a radiator fan?

Keith Tanner
Keith Tanner GRM+ Memberand MegaDork
11/13/18 2:37 p.m.

Psst! Check out steps 8 and 9.

I understand the urge to reuse off-the-shelf parts as much as possible - but I would not have used pusher fans. It's not like 40 year old MGs are renowned for their awesome cooling systems anyhow. I would have put a big puller on the backside with a shroud. There's loads of space and it would be a lot more effective.

dculberson UltimaDork
11/13/18 3:04 p.m.

Those hammer formed mounts are beautiful.

Keith Tanner
Keith Tanner GRM+ Memberand MegaDork
11/13/18 3:28 p.m.

You can always count on brackets that are much prettier than is strictly necessary from Carl & co.

OnTheChip New Reader
11/13/18 3:32 p.m.

How did you determine the height for the expansion tank to ensure a complete fill such that the water jackets in the head have no air pockets?


mazdeuce - Seth
mazdeuce - Seth Mod Squad
11/13/18 3:50 p.m.

I'm slightly disappointed that you didn't round off the corners on the bracket in step #17. 

Carl Heideman
Carl Heideman
11/13/18 4:02 p.m.

Thanks for the kind words.

Regarding the fans, I agree about puller fans 99% of the time.  However, the stock 77-80 pusher fans really blow a lot more air than almost every aftermarket fan we see.  We regularly remove puller fans from 77-80 MGBs and put them back to stock to solve cooling problems.

Regarding the expansion tank, we just want its inlet/outlet up high enough to bleed air from the radiator and honestly just go with the MG factory locations.

77-80 MGB radiator setups were carryovers from the MGB GT V8s, so they're very robust and we have found them to work well in a lot of engine swaps.  I've had this setup on my Miata-powered MGB since the mid 1990s.  

Of course, this swap is our first one with AC and a supercharger, so we may be testing the limits...if so, there will be a 2nd installment.

p.s.  Since the story was written, the corners from step 17 were rounded off--good catch.  I  try to radius every corner because I'm not a fan of unnecessary bleeding...

Keith Tanner
Keith Tanner GRM+ Memberand MegaDork
11/13/18 4:29 p.m.

I'll be interested to see how it does for cooling. Supercharging and AC are a double whammy that's going to work that system harder. Those MG fans do look like they have big, high-torque motors which is better than most of the "slim line" units on the market. If you do go to a puller, look for the biggest motor you can find. We have some good Spal part numbers if you're looking.

noddaz GRM+ Memberand SuperDork
11/13/18 6:37 p.m.

What about the water nipple on the radiator that is pointing the wrong direction from the expansion tank?

And thank for this type of write up!  It is most helpful for a newbie like myself to see all the little things that make for a great swap.

Carl Heideman
Carl Heideman
11/14/18 7:52 a.m.

That radiator picture is actually a mockup from before we sent it to the radiator shop.  We had the nipple turned the correct direction while they had the radiator apart.

Regarding how this will hold up, it's a restomod street car that will get some aggressive driving in short bursts, so I'm fairly confident it will be okay.  This setup will probably struggle on a track with extended WOT/high horsepower periods of use.

For me, radiators are very frustrating because they tend to be very trial and error.  If there was better math published and understood, sizing would be so much easier.   For example, I think they should be rated with a horsepower/duty cycle with some sort of multiplier for air flow.  Start with the Otto cylce rule of thumb:  1/3 of the engine's power (heat) goes to the flywheel, 1/3 goes out the exhaust, and 1/3 goes into the radiator.   At idle, an engine is making maybe 2-5HP no matter how big it is.  All it's doing is spinning the flywheel and accessories and it doesn't take much power to do that.  At WOT and high RPM, the engine is making its peak power.  What really matters to the radiator at WOT is how long the engine is at that peak range of power.  On the street, it's in bursts of a few seconds at a time.  On a track, it's much longer periods of time, almost continuous, for maybe 20 minutes.  Whenever it's making that peak power, the radiator has to absorb it.  That's where the duty cycle comes in.  What percentage of a period of time is the engine at peak power?  Then you bring air flow in:  At idle, the engine isn't making much power so the radiator isn't working very hard, but it's got to have air going through it.  Most idle cooling problems are air flow, most peak power cooling problems are radiator sizing (sometimes with air flow as a factor).  

Anyway, that's what's going on, but there isn't much useful information/math to help us with sizing.  Frustrating.

By the way, some members of Hope College's FSAE team are working on just this math for their car--should be very interesting.

You'll need to log in to post.

Our Preferred Partners