The "Pink Milkshake" - How worried should you be? 200k Radiator Teardown

Should you be worried about the dreaded "Pink Milkshake"? I seek to provide some insight.

SUMMARY / TL;DR: I tore apart my original 200k miles radiator for signs of failure that would lead to the Pink Milkshake. I could not find any potential failures. Although a residue was present on the bottom of the radiator, there was no signs of leaking, rusting, on the cooler. A far as I can tell, the radiator was perfectly fine and probably would have last much longer with regular coolant and trans fluid changes.


BACKGROUND: I bought a 3rd gen 4Runner last year with about 192k miles on it. While I was researching the "what to look out for", one of the biggest concerns was the dreaded "Pink Milkshake". The Pink Milkshake is the mixing of the automatic transmission fluid with the coolant/antifreeze, resulting in a creamy and frothy looking pink liquid that circulates thru the transmission and depending on when it is caught may require a transmission replacement.

My question was, how common is this issue? Is this a serious worry? The forums have a way of biasing us, we typically hear the "bad" more than the good. If someone has the issue, they are way more likely to take to the forums than the majority of folks whose radiator is perfectly fine. We typically pay attention to the bad more than the good. If you spend too much reading the forums like I do, you probably run into these general schools of thought when it comes to the preventing the Pink Milkshake.

1. Regular coolant and transmission fluid changes is enough to prevent.
2. Replace the Radiator every a certain interval (every timing belt change, 100k-150k miles, etc.) for peace of mind.
3. Separate the OEM transmission cooler (external trans cooler) and bypass the OEM radiator all together and therefore no chance of the fluids mixing.
4. The OEM cooler sucks and is not enough to cool. Bypass that tiny OEM cooler with large external cooler or put another in series with the OEM cooler.

My T4R is now at 200k miles and overdue for a timing belt job. Before the teardown, I aligned with school of thought number 2: Replace radiator at timing belt change. This is the original radiator that the vehicle was installed with and never replaced. According to the previous owners maintenance records, the coolant and trans fluid was changed before, albeit not at the Toyota recommended interval, but non the less changed a few times in its life.

Now that the timing belt/radiator replacement job is done, I decided to completely disassemble the original radiator, to see if I can find any signs of failure, inspect the overall condition, learn something, and perhaps guide my future decisions on radiator replacement. Come along for the ride.


TEARDOWN:



I only removed the bottom head/cap of the radiator (black plastic). The bottom head is the radiator outlet (cooler side) where the coolant leaves the radiator. The top head is where the coolant inlet (hot side) enters the radiator from the engine. The coolant has already cooled significantly by the time it reaches the bottom. This is why the bottom contains the transmission cooler (heat exchanger would be more technical, but I'll refer to it as a cooler). The head is pressed on with a large perimeter o-ring (more of a square than an "o") that seals the coolant between the head and the radiator. The head is pushed into place and tabs along the perimeter of the radiator hold it together semi-permanently.




With the head removed you can now see the transmission cooler. I don't know what I was expecting, but I thought it would be... bigger? It's just a hollow tube, approximately an inch in diameter. The transmission fluid flows in the space between two thin tubes that was capped off and brazed. The fittings themselves are blazed into each end of the tube shell. The coolant flows through the middle of the tube and around it to exchange heat. The heat exchanger cools the transmission when the transmission is hotter than the coolant. It also warms up the transmission when it is cold (at startup), as the engine tends to warm up faster. This is the advantage of having the trans cooler in the radiator, the trans fluid gets to operation range much quicker.




I was a bit alarmed to find some pinkish sludge at bottom of the radiator head, right by the trans cooler fittings. My initial thought was a small leak had formed in the trans cooler, which mixed with the coolant. However, no oil was noted in the coolant prior to this, and the sludge wasn't oily, mostly slimy and gritty. You can see a comparison of the oil vs the sludge.



Just to be sure, I conducted a leak test. I pressurized the trans cooler with 30psi of compressed air and sprayed the whole thing with leak check fluid (not just soapy water, the good stuff). No leaks were found. Phew.









I removed the thin nuts that were holding the trans cooler to the head. They were rust seized so I used an impact gun on them. Unfortunately the force of the impact slightly twisted the cooler that you can see in the photos after it is apart. I was surprised to find it still held pressure despite the twisting. Once I had the cooler out, you can see the sludge buildup. It felt grainy, almost like sand. Perhaps from various casting that are on the engine? I'm not too concerned at this point.






You can see how cooler seals with the radiator (see sketch for reference). There is a small o-ring that sits inside the radiator that seals the cooler against the inside wall of the radiator head. On the outside, the thin nut bears down on a washer and a rubber gasket against the outside wall of the radiator head. This would not seal, as any liquid could potentially pass thru still via the threads of the nut. I think the rubber gasket was added as a "isolator", acts as a cushion to prevent the metal washer from bearing directly on the plastic, and adds some debris/water resistance to the fitting.



I probed the trans cooler with a magnet, the fittings are some sort of steel, whereas the cooler tube is non-magnetic, perhaps stainless steel as some grades are non magnetic. I did a quick grinder test on the fitting and the tube material, both produce sparks (aluminum does not).





I cut open the cooler to see what was inside and inspect for corrosion. It was interesting to see a metal mesh inside the space of the two tubes to aid in the heat transfer from the fluid to the tube shell. No signs of corrosion were found inside the tube.


QUICK DISCLAIMER: This is in no way a scientific study, more of an inspection. This is only one data point and not enough samples were taken to draw an reasonable conclusion across all vehicles still in service.


THOUGHTS: I was expecting to see some sort of degradation of the cooler, and surprised to find nothing that would lead to a mixing of the fluids. I was surprised to see how small the exchanger was. The only signs of corrosion were of the outside of the radiator, were the send and return line fittings thread into the cooler. The trans cooler is pretty well contained, and would only mix with the coolant if the portion of cooler inside the radiator fails (ie, the tube, brazed joints, material corrosion).

I have heard in the past that an "o-ring" is the only thing that separates the trans fluid from the coolant. This is not true. If the o-ring fails, coolant leaks to the outside of the radiator, and does not mix with the transmission fluid. Theoretically, if the fitting on the outside failed in just the right area, I could see the o-ring becomes the only thing holding the fluid back from mixing with the coolant. If that is the case, I would expect to see some trans fluid leaking to the outside of the radiator as well. See sketch below:




CONCLUSION: Based on my findings, I am much less concerned about the Pink Milkshake. The 200k mile cooler was nowhere near failure. If I had to guess, the failures reported before where either due to high levels of salt intrusion, poor maintenance causing acidic coolant or trans fluid that degraded the cooler, or tiny portion of manufacturing defects which lead to premature failure.

Personally, I will not be replacing the radiator at the each timing belt change. Unless I see significant exterior corrosion on the transmission cooler fittings, damage to the fittings, or trouble signs in my coolant (oil in coolant, discoloration, etc). With regular coolant and trans fluid changes, I see the Pink Milkshake risk as very small.


Thanks for reading. Constructive criticism welcome.

***NOTE: I have this writeup cross-posted in TacomaWorld.com, as I am more active there.

 

Nice. This is really cool. My OEM radiator failed by leaking coolant to the outside about midway up the radiator. The OEM failed at 125k miles and the Advanced Auto replacement failed at 250k miles in a similar way. Halfway up on the driver's side.

Nice to know that the transmission cooler on a 2001/2002 isn't just separated by plastic.

I wonder if from 96-2001, was there any revision in the radiator design? This may be your answer as to why you don't see any reason to worry on your original OEM one. Earlier radiators may, indeed, be prone to failure.

I wonder if the Advance Auto one I have is built as well as the OEM. I think I am going to just go to a 100% metal radiator next time.

 

What would be over-the-top cool, is if you can get the next poor soul that gets the pink milkshake to send you their radiator so you can redo your analysis on that one.

 

It's a good question. I'm not sure if it was redesigned. There appear to be two separate part numbers for the radiator on 3rd gens with the 3.4L 5VZ-FE that fit my 2001 according to the Toyota Parts Website: 16400-75181 & 16410-AZ011. However, if I select a 1996 4Runner, I am only given the 16410-AZ011.



As far as I understand, the only difference between the two is the country of manufacture. One is a "cheaper" alternative made in Taiwan to the Japanese made one. At least that is what I gather from this thread on T4R.org about it (below). It could certainly be that the cheaper one is made worse, however I cannot confirm that without further datapoints. I haven't come across much info from people who've actually had the pink milkshake to make any sort of determination on that.

Agreed. I'd be willing to do that.

 

Well, an old revision would be long gone. Looking at current available part numbers isn't really going to tell you much. You'd have to collect old junkyard radiators and look for engineering numbers/revision numbers.

The same part number often spans many engineering\revision numbers. Only when you have a major part revision do they put out a new part number. Or when they re-engineer a part to work across more platforms so they can supersede a few old parts.

 

Do you see a revision number or part iteration number on the radiator? I'll check it when I get home for any clues. Is there a specific format to the numbers?

 

For TOYOTA, I'm not sure.

TOYOTA may not even put many numbers on parts. I do know that the 2001-2002 Throttle Body has the part number stamped right into the top and it's a good part number, not an engineering number. Don't know on the radiators. My old one is long gone.

I would look the radiator over and see. To me, it would make sense to stamp something on there somewhere to be able to ID the thing. A revision number of some kind helps a car manufacturer know what revisions are failing. Maybe TOYOTA doesn't do running changes like the American car companies do so they just track Car VINS and can map it back.

I've dealt with restoring Fords and used to be deep in the weeds of part numbers and revisions. For Fords, the number stamped or printed on the part itself was very close to a part number. The prefix would tell you the year of revision and the middle of the number would be a combination of the area it went to on the car and the part. The suffix was usually the revision. We called these engineering numbers and except for very few, you could not cross reference them to actual part numbers the parts guy would understand. For a part assembly that had a single part number, you could find more than one of these engineering numbers on the actual individual parts that made it up.

 

I bought a 2001Limited with 344k miles a couple months ago. Drivetrain seems to be very well taken care of, but I was worried about the stock trans cooler. I had planned to bypass it and put an external cooler on today. This thread is making me rethink that. I don'tike the thought of the transmission not getting warm enough anymore than I like the thought of pink milkshake.

If it helps your part number discussion, here is a pic of the visible markings on the top of my radiator. I have to think this is a replacement. No way it's original at 344k.

 

Looking up the radiator part number for a radiator gives: 1640075181

TOYOTA uses the first 5 digits for the department/area of the car (16400). They didn't print it on the radiator since it's obvious it's a radiator and part of the cooling system.
75181 is the main part number. Dunno why the 1 is missing off the end.

422171-1844, if I had to guess looks like an engineering number/revision number to me. I was able to find pictures of identical radiators but with just the last digit different.

2C may be a date code. 2C sounds like 2002 March...but doesn't make much sense if you have a 2001 car. It may be a lot code instead.

Honestly I bet its original to the car. Very few people opt to replace a radiator with a $600+ one from TOYOTA.

PA66-GF25 is the plastic used. PA66 = Nylon 66, GF25 = 25% Fiber Glass Reinforced.

 

It's not the plastic that fails, per se. It's where the plastic is crimped down, both top and bottom.
I've seen threads on how-to tighten the crimps down to prevent the leaks they develop. Looks pretty easy to do.

Personally, I put 3 core, all metal, radiators in both my 87's. The 4Runner's failed a couple years ago. Blew it's top off.
The one in the pickup is still going strong, though. Since they both lasted more than 25 years, so I think it was a small investment that has paid off big so far.
Come the time the radiator, which the PO, a dealership, put in my 99, along with a new timing belt, water pump, etc, just before I bought it, needs replacement, it's going to get an all metal 3 core radiator. Probably CSF, but we'll see...

Have fun all!
Pat☺

 

Are you talking about the crimp around the outside of the radiator holding the bottom and top plastic reservoirs on?

A failure at the crimp isn't going to cause a pink milkshake...which is what OP was investigating.

What I didn't realize is the transmission fluid runs through a metal tube like that.

 

If you still want to do an external cooler, you could add a thermal bypass if you want your transmission to heat up quickly. I've heard some folks add a thermal bypass valve to their cooler lines. Basically, the thermal bypass is a valve that opens at a certain temperature. When it is cold, the fluid bypasses the cooler and just recirculates back to the transmission thereby heating up faster. When it reaches the "open" temperature, it opens to allow the fluid to flow to the cooler instead and closes the bypass passage off. It probably won't heat up as quickly as having it in the radiator, but it is a good compromise solution.

 

I keep buying manual transmission vehicles. I wasn't thinking auto tranny problems. Sorry.
Pat☺

 

I have the same part number. And I have owned this T4R since 2004 and it has never been replaced nor has it leaked, nor has it ever overheated. 335K miles and counting.

 

I would agree. I have the same numbers on mine and have owned the truck since 2004@30K miles. It is now at 335K miles and has never had any issues. We have done 2 tranny flushes, 1@ 150K miles and 1@ 300K miles. No issues.

 

CattyWampus: Thanks for the in depth testing and tear down. Awesome information. As I mentioned in some of my other replies. Ours seems to be the original radiator and at 335K miles we have had zero issues. Tranny flushed at 150K miles and 1 @ 300K miles. Never has she overheated, leaked or ? Never have seen the dreaded "Milkshake" horror.

 

Yes, very interesting analysis. Thanks for posting.

I opted to have my radiator, among other things, replaced while my mechanic was performing the t-belt service. I believe my radiator was original to the truck with 188,500 miles at time of replacement. The Japanese made version is no longer available from Toyota, but the Taiwanese version is. At $108 from the dealer, it's cheap insurance.

 

thanks for the info. i agree about cheap insurance, i bought a new radiator to install when i install the water pump slash thermostat slash timing belt. i also bought a Hayes 699 trans cooler, and i am going to bypass the radiator. how Cool is that?

i am sensitive about temperatures. when i was running the truck around the yard to get temps warmed up, water peaked and held around 186º, and trans peaked at 204º, which is too high on a 60º ambient day. suggests the pump has retired. i'm getting a 170º thermostat for the new pump.

for me the question about the risk of the milkshake isn't satisfied by the record of not failing, it is concerned with the opportunity to fail. I can mitigate that with an oil cooler that guarantees that is will not happen, for less than a hundred bucks. Cheap insurance indeed.

Ciao Runners