OhSix, I love your dedication to this. I think we're all learning a lot from your time out in your garage!

Repeating this process and picture series on the Spectra radiator, with the copper fittings and exchanger, would really make for an authoritative study on the mechanics of the problem and advantages of an alternative design.

 

Thanks Mr. Hokiefyd. Glad to share - although pretty sure my OCD is showing.

I'm so disappointed in OSC, I've actually toyed with the idea of cutting this brand new rad up just to see whats inside. If it weren't Christmas time, maybe...

I'm pretty amazed the OEM rad holds up as well as it does. The idea that aluminum, plastic and synthetic rubber are holding an assembly together and keeping fluids isolated over many thousands of miles, heat cycles in all kinds of environments - is something to ponder.

OK, so these are just gratuitous images but they're cool.

 

Impressive work OhSix! I like it.

 

I wonder if all aluminum stacked plate exchangers (including the OSC) would have this type of connection? The Spectra made it clear that if you have a copper fitting, then it's likely male studs braised straight to the concentric tube exchanger, but it may be that none of the aluminum ones are that way, and all of the aluminum ones use the female threads with o-ring.

 

schwejo's images caused a curiosity itch that just had to be scratched. So, the dice was rolled and a trans port was removed from the new OSC rad - which uses a similar two stage assembly at the OTW exchanger.

here's what was seen:

Conclusion:
The two nut design indicates two discrete mechanisms. One attaching the OTW exchanger to the tank with MALE threaded ports protruding thru the tank. MALE being specific to the OSC rad.

The angled tube is similar to the Spectra in the sense that it is terminated by a flare compressing to the port extending from the OTW inside the tank. Same design used in household plumbing.

Although the image is difficult to see, there is both a spring steel "Belleville" washer and stainless spread washer under the threaded shaft extending thru the tank. This implies corrosion is still possible but in this design, BUT expansion caused by electrolysis in that washer will actually pull the seal tighter between the tank walls and the external nut responsible for compressing the OTW to the walls.

Peering into the opening, it isn't clear that the OTW is staked plate, however, based on every image of other OSC rads, its a safe assumption it is.

This design is a hybrid of the Denso built OEM design. The OSC (and similar designs) employs a "fail safe" for fluid isolation. Instead of relying on the port to provide BOTH mechanical attachment and fluid isolation, the tasks are spread to two components.

Putting it all together:
- The OSC has only slightly fewer rows in the core than the Denso, MANY more than the Spectra (although the Spectra uses wider individual rows and ports at the upper and lower braised plates where tanks are crimped into place).
- The OSC uses staked plate aluminum OTW similar to the Denso, but offer FAR superior fluid isolation and long term mechanical integrity.
- Both the Spectra and OSC exhibit "waves" in the Z axis of the core and side plates, where a (near) 10 year old Denso was "perfectly" straight and visually properly aligned. What does that mean? IMO: more repeatable mechanical assembly in the manufacturing process - which translates to improved assembly accuracy, higher first pass yields in production and "should" mean longer service life in the field. HOWEVER, any long term benefits to that are offset but an inferior design responsible for isolating incompatible fluids (meaning SMOD catastrophes at some point if not closely monitored).
- The Koyo *seems* to employ the same two part design as the Spectra and OSC, meaning they have separate mechanisms to attach the OTW exchanger and angled ports to conduct trans fluid. The difference in design being: the Koyo exchanger has a female threaded exchanger port where the OSC has male. In this regard, the Koyo duplicates the Denso built OEM design (with the inherent limitation of few threads) where the OSC duplicates the larger threaded shaft (and potentially more robust nut) design seen across many aftermarket rads. IMO: both designs are superior to OEM based on one fact: it would take two (seemingly unrelated) failures to allow SMOD, where any mechanical failure in the OEM design guarantees SMOD will happen.

 

This design is a hybrid of the Denso built OEM design. The OSC (and similar designs) employs a "fail safe" for fluid isolation.

Are you sure it does? If the aluminum exchanger has female threads, and the male fitting uses an o-ring, then that fluid connection is still internal to the tank. I don't see the significant difference, here, between this and the OEM design. If anything, there's an additional connection (the second flare nut on each fitting) that might could be avoided by using a one-piece fitting like on the OEM radiator. The Koyo design uses the female threads of the exchanger to hold it tight to the tank, just as the OEM design does. True, the actual ATF pipe elbow attaches separately, but the exchanger connection, mechanically, is similar to the OEM design.

The Spectra radiator (and I assume the TYC also, and anything with brass fittings) uses no o-rings, with a male threaded stud braised or soldered to the concentric tube exchanger. Assuming for a minute that braised connection is solid, that moves the opening of the exchanger to outside of the tank. But an exchanger with internal female threads is, to me, replicating the OEM design. The o-ring is the indicator of this -- they have to make that connection fluid-tight.

The only major difference I see is the o-ring is on the coolant side of the threads, vs. the OEM design where the o-ring is on the ATF side of the threads. I'm not sure that's significant, since that SHOULD be a dry area, anyway (because of of the larger o-ring seal to the tank).

 

Very Nice analysis.

 

OhSix, I'm saying, if you separate the pipe elbow and flare nut from the Koyo's exchanger fitting, and install just that fitting alone, it would likely look identical to the OSC. In other words, it's my guess that the OSC uses the same type of stacked plate exchanger that both the OEM and Koyo use -- with female threads and o-ring seal at that connection (in addition to the larger o-ring to seal the exchanger to the tank).

I'm not sure I've ever seen an aluminum stacked plate heat exchanger with male threaded studs. As I browse images on the internet, I can't find a production-type (IE not high performance or racing application) stacked plate cooler that doesn't use female threads. This page shows the concept, and it looks remarkably similar to the Koyo:

http://icemancooling.com/category/engine-cooling/radiator/

This is about halfway down that page. The staked plate exchanger has female threads, into which the fitting passes through the tank and then fastens.

Another one:

http://www.dana.com/light-vehicle/p...ucts/engine-technologies/engine-oil-cooling/radiator-in-tank-engine-oil-coolers

Both the OE and Koyo's aluminum plate exchanges use female threads, and I would be surprised if the OSC's exchanger didn't as well. I could be wrong. I suppose the only way to know for sure is to verify that the large nut on the OSC's fitting really is a NUT, and not part of the fitting itself, like the Koyo. If the OSC's nut really IS a nut, then I think that's probably the best setup -- a stacked plate cooler with male threaded studs welded to each end. That would surprise me, though. Knowing that these are commodity-type parts, it's likely that OSC uses the same suppliers of these exchangers that others use. We can already see the similarities in the OSC's and TYC's cooling stack (with the fins ending a half inch from the bottom).

 

Ah! Finally got your meaning.
Big fat DUH on my part. And big fat assumption (mine) that the OSC in male too.

Your point aligns with images from the OSC pdf file. Chart update eliminating my assumption error in progress.

Good catch!

Corrections made. See post #50. Thanks Hokie

 

Are you comfortable drawing a conclusion at this point in terms of radiator rankings?

 

It appears we have enough information to derive rankings based in mechanical design/construction. No doubt assigning a rank would open up further discussion on ratings I might assign - but that's never stopped me from having strong (& erroneous) opinions. LOL.

After dissecting the OEM and Spectra rads, it occurs to me:

The stacked plate OTW design is a multi-phased, multi-element assembly comprised of tight tolerance sub-components requiring precision manufacturing techniques. Creating each hollow plate is - a complex procedure demanding precision material formation and several assembly steps on its way to becoming something you could hold in your hand. In my imagination, this component is entirely (or nearly entirely) machine built - because it's not possible to build such a component with the inevitable variability introduced by humans. The expense involved in designing & producing purpose built machines to make this component is no small consideration. Add to that configuring a factory with those machines, designing process controls, inspection criteria, supply chain material management, many other mass production controls along with demanding QA steps required to achieve maximum yield and you have world class manufacturing of a small but critical component. While Denso may have been the selected supplier for the Honda design, its safe bet the OTW in our radiators was sourced from the vendor hokie linked to in this thread. That supplier might be "lowest bidder", but the only reason they could be is the might literally "own" this segment of automotive components, selling to nearly every OEM auto manufacturer and aftermarket producer, which likely brings their COG and COP down while meeting every quality standard out there. The point is, the stacked plate OTW exchanger just may be the most expensive sub-assembly in the RL rad.

Compare that to the Spectra. While it may be the most robust design in terms of isolating fluids and physical mounting to the tank, it is a pretty simple assembly to construct, consisting of two tubes of different diameters, one inside the other. The space between those tubes is wrapped with thin sheet of brass, stamped in a pattern creating a varied surface for the fluid to traverse thru. At the outer edges of the larger tube, male fittings are braised into place, the ends of the tubes are crimped and braised (likely at the same time/in the same braising oven as the fittings). While there may be machines involved in production, its literally nothing compared to the complexity of what we see in stacked plates. The concentric design lends itself nicely to the cottage industry producers of the world. In that segment, lowest bidder means something else than the builder of something as complex as stacked plates. While there is a fair amount of assumption in all this, it should be obvious the Spectra is cheaper to manufacture than other designs. Still, in the absence of manufacturing errors creating pin holes (ETC) looking only at the mechanical interface, I'm attracted to this design.

The missing data is heat management effectiveness. With the variability inherent in temperature measurement techniques, this would be a tough nut to crack. Other than sensors sunk in fluid at each point (or device) in the flow of heat management components, we have a measurement conundrum that may never be solved across the different producers of radiators. To derive a meaningful comparison between OTW, OTA and WTA exchangers seen in the aftermarket, we need repeatable methods to sense and record temp data, but we simply don't have that. However, I do have a way to get some data - at least to compare OEM to the OSC. That data should be available in January.

Ranking rads still requires some assumptions. In that regard, I assume Honda has implemented the *best* all around heat management schema. That's a pretty safe assumption because - after all - they are privy to a long history of R & D test results that (no doubt) has netted mathematic models and design criteria on which to protect their consumer products. The aftermarket has similar models but they simply cannot design to the same level as an OEM. As we know know, the single biggest issue with Honda's implementation is the weakness of there fluid isolation mechanism.

So... here's my stab at ranking. For fluid isolation, I am assuming the two stage mechanical designs offered by KOYO and OSC in these examples lower (or nearly eliminate) the potential for a SMOD event making them far superior to the OEM implementation. Although I lean to the single structural design of the Spectra, the reliability delta between it vs KOYO & OSC might be hair splitting. In terms of absolute mechanical isolation, I think Spectra wins in that category. (except in the potential for leaky-ness, where the stacked plate wins due to superior manufacturing)

This has been an instructive learning experience. And there's more to come in the next month or so. Special thanks to hokiefyd who caught errors in logic and oversight of evidence, schwejo who offered insight to the KOYO and carsmak who donated Spectra and OEM rads for dissection.

Edit # 26: changed chart & message body syntax

 

In addition to what you said, OhSix, I believe the heat exchanger thermal capacity issue of the Spectra to be diminished to some degree by the Ridgeline's standard external air cooler. I used the Spectra type radiator (TYC) in my MDX, and I'm glad that I added the external air cooler to supplement the brass concentric tube cooler. I do think it probably has less thermal capacity than the stacked plate aluminum cooler. I think it'd be good practice for anyone using a radiator with a brass exchanger to use an external air cooler as well -- just to be safe. In other words, I think the aluminum stacked plate exchanger has more capacity, but it's something that can be mitigated to some degree.

I think the Koyo and OSC design is similar to the OEM design in terms of fluid isolation. Like the OEM design, it uses the female threads of the exchanger to fasten it to the tank, and it uses a small rubber o-ring for protection. The only difference I can perceive, in terms of this connection, is the Koyo design uses an o-ring on the coolant side of these female threads, whereas the OEM design used an o-ring on the ATF side of these female threads. Whether or not that is significant, I do not know.

I'll bring up here the thoughts of a friend of mine, who is a mechanical engineer at an automotive supplier in Michigan (though not a cooling systems supplier). In summary, he believes that there is a galvanic corrosion process, here, and it's made possible by the one dissimilar metal in the entire connection (in the OEM design) -- the ONE stainless steel washer. Because that stainless steel washer has a different anodic index than the aluminum, the aluminum becomes an anode and begins to corrode. The threads have the least surface area, so they suffer the worst. The presence of acidic coolant, as an electrolyte, strengths the "battery" process, there. He believes that if Honda had used an aluminum washer, instead of the stainless steel, this problem wouldn't be nearly as pervasive.

I asked him about the brass exchanger -- brass has a much different anodic index than aluminum as well. I'm thinking about possible corrosion of the main aluminum cooling stack. He said that dissimilar metals should never be used when it can be avoided. Interestingly, the Spectra offered for post-mortem, here, had a leak in the aluminum cooling stack, right? I wonder if there was an electrolytic process going on with that one, or if it was just a manufacturing defect that allowed that pinhole leak to form after only a few years.

My friend's comment, either way, was that the more acidic the coolant becomes, the better it's able to support and sustain a galvanic corrosion scenario. I think I may go to doing a simple drain/fill of the radiator once a year. It's only 25 bucks for a gallon of coolant. Sure, after five years of doing that, you've bought a new radiator, but I'd much rather spend $125 doing simple drain/fills than replacing radiators.

Assuming that theory is correct, I don't think any of these radiators stands head and shoulders above the rest. The OEM Denso/Koyo/OSC have the better heat exchanger, and the only dissimilar metal involved is potentially a stainless steel washer. If one could remove that washer and replace it with an aluminum washer (such as a crush washer for the oil drain plug), then you'd think that the potential for this problem recurring is diminished greatly. The Spectra/TYC radiators, on the other hand, remove the problem of fluid mixing through the mechanical connection (since that connection has been moved outside the tank), but you still have to consider a manufacturing defect of the internal exchanger (a pin hole, which would be particularly dangerous because you could have fluid mixing and be unaware of it), and you still have a potential issue of dissimilar metals. You wouldn't want to move the problem to the aluminum cooling stack, and have the fatigue move there.

All of this is probably picking nits at this point, but I'm sure it's all similar discussion to what happens in conference rooms at OEMs and suppliers all the time. What set of compromises do we take, and what risks are we happy with? I think we can all agree that the best solution is a 100% aluminum radiator with a non-removable heat exchanger, welded to the lower tank (so no possibility of a leak), with nose connection nipples welded to the lower tank (so no threaded connections at all).

But that'd cost $500.

Maybe it'd be worth it...

 

In addition to what you said, OhSix, I believe the heat exchanger thermal capacity issue of the Spectra to be diminished to some degree by the Ridgeline's standard external air cooler. I used the Spectra type radiator (TYC) in my MDX, and I'm glad that I added the external air cooler to supplement the brass concentric tube cooler. I do think it probably has less thermal capacity than the stacked plate aluminum cooler. I think it'd be good practice for anyone using a radiator with a brass exchanger to use an external air cooler as well -- just to be safe. In other words, I think the aluminum stacked plate exchanger has more capacity, but it's something that can be mitigated to some degree.

I think the Koyo and OSC design is similar to the OEM design in terms of fluid isolation. Like the OEM design, it uses the female threads of the exchanger to fasten it to the tank, and it uses a small rubber o-ring for protection. The only difference I can perceive, in terms of this connection, is the Koyo design uses an o-ring on the coolant side of these female threads, whereas the OEM design used an o-ring on the ATF side of these female threads. Whether or not that is significant, I do not know.

-snipped by OhSix-

But that'd cost $500.

Maybe it'd be worth it...

Thanks so much for the thoughtful input and information from your engineer bud. Always good to have multiple perspectives on the threats to safety and reliability of failure modes like these.

You are correct, carsmak's donated Spectra rad had an unidentified coolant leak presumed to be at the lower plates of the core.

Now that the grunge of opinion has (hopefully) been washed from my cranial cavity, I think the point you've patiently been making is clear. Allow me to restate just to ensure I've got it.

The naturally occurring process of galvanic corrosion is going to happen wherever dissimilar metals are present. Due to that fact, any radiator design employing dissimilar metals will suffer corrosion simply due to their presence in mechanical construction - which is true of every sample we've seen so far.

In terms of fluid isolation, the only real stand out among these samples is the Spectra, because: a mechanical failure related to its OTW exchanger in this design will almost certainly prevent fluids from mixing - at least - to a much smaller degree than the OEM or OSC or KOYO - all of which employ similar mechanical attributes of a male/female AL fitting to the OTW exchanger.

So... we have two factors working against us. Internally generated corrosive processes which are exacerbated by external environmental exposure.
Do I have that right?

If so, we have little if any difference in design when looking at the OEM/OSC/KOYO. The only real differentiator is the slight improvement in mechanical offered by the KOYO and OSC over OEM - where the task of attaching the OTW exchanger to the tank is and conducting trans fluid distributed to two parts. But both are susceptible to internal and external corrosive effects.

One of the things that tripped up my thought processes were (what I thought was) lack of evidence of externally visible corrosion in many images offered by forum members. I see now that I was linking those images to support what I thought I knew about this topic - I kept thinking: no external evidence = galvanic corrosion is a nothing burger for those owners in mild climates. PFFFTTTT. wrong again.

I *think* I've turned another corner on this topic. Finally come around to your well formed investigation and experience. Thanks so much for continuing to educate! Much appreciated.
The only complaint now is: I gotta revise previous posts to reflect reality. Small price to pay for an improved foundation of knowledge.

FYI: I been in contact with two race radiator manufacturers. One specializes in hot rod applications, the other produces both OEM style replacements and "race" applications. The former estimates $600 to build an all AL, the latter estimates $450. The casual inquiry didn't include details around the OTW exchnager, so there's still a ? mark on that. Having just passed 116K on the clock, not sure I'd invest, but the idea is in the hopper just in case...

 

Quick question after scanning the above excellent info, but how does acidic coolant affect the washer as it should never be in contact with it? Or did I somehow miss the point?

 

In electrolysis, dissimilar metals don't need to be in contact with the electrolyte solution generating current flow. The metals only need to be "in the circuit". In that context, trans and engine coolant are the electrolyte passing thru AL. The spring steel washer (Belleville - sp?) is in the path of AL components, so the "charge" passing between the metal creates the electrolytic induced galvanic effect.

WHEW. I'm exhausted just typing that stuff. :act006:

 

Ah, I see.

Regarding replacing the Belleville washer with a sealing washer as mentioned above... seems it would defeat the purpose of the Bellevile washer... which is to put a designed tension on the connection... as I understand it.

"Also known as conical washers, Belleville washers are designed to maintain tension in bolted assemblies with spring-like action. This creates a tight hold and uniform pressure that remains constant through heat and vibration, which might shake standard washers loose."

https://www.fastenal.com/products/f...s"|~ ~|categoryl2:"600089 Washers"|~ ~|categoryl3:"600090 Belleville Washers"|~

 

Far as I know (which doesn't go very far) there are several ways to achieve tension attributes in washer designs. From simple splits to the variations in "teeth" like inner, outer and nord lock, they "could" serve as compression for predictable assembly as well as tension (or) "locks" against vibration. I'm positive Honda knows the risks and benefits of all of these, just as they are electrolysis and everything else we are discussing. I'm equally positive the convex conical was chosen for reasons of both factory assembly torque/port clocking as well as tension to retain clock orientation under vibration and heat cycle models used in product development.

Why did the aftermarket follow Honda with spring and stainless washers? Beats the heck out of me. If they are going to follow an existing design, why not build their radiators with the single port/clamp design? I dunno.

Honda made their design choice for reasons other than absolute structural integrity (read: good enough/CYA/minimal warranty exposure).

With that said, an "easy" fix might be changing out the conical & stainless washers to AL - or an electrically inert material like composite fiber - virtually eliminating concern over electrolysis. Should a DIY enthusiast decide to take that path, a critical things to consider is stack up height in the washer(s) and torque. Stack up being critical to complete thread engagement and torque being critical to both tension and prevention of thread strip. After seeing - up close and personal - the female port in the OTW having 4 threads to keep its poop in a group, doing any sort of disassembly to a new radiator would be a total crap shoot IMO.

 

Speed, I think we can apply for engineering credits once we complete this course.

 

It has been said many times that the failure occurs from the inside out, but what's the specific failure mode in that scenario? Could just be symantecs and perhaps I've used "deterioration of O rings" to loosely, but assuming no outside indication, what specifically causes the breach of the O rings, especially the smaller one that seals the ATF connection. Seems something must pull the smaller O ring from its landing and atleast in my mind, that presumes some sort of external signs, unless the cooler is falling away inside the lower tank, but from pics I still seen, that was caused by swelling of a decomposing Belleville washer. Or so it seems to me.

No doubt galvanic corrosion could be the root cause, but given proper use of industry standards materials, the galvanic corrosion may be slow enough to be negligible for the reasonable life of the car and on par with the so many other items that can fail on a car. That's my hope atleast with the Koyo I installed. It's still interesting that my 06 Ridgeline started showing outward signs while my friends Pilot that is several years older and "lives" in a similar environment shows no signs of corrosion. I'm still hanging my hat on an improper Belleville washer while agreeing that galvanic corrosion may be the cause but be negligible when proper materials are used. My apologies for pecking this all out on a smartphone....I'm at the limits of my editing ability!

 

What's confusing to me is why Honda apparently used both a stainless steel washer AND a bare nekkid steel washer in the connection. Both connections on my MDX had one pretty good-looking washer and one pretty tore-up looking washer. If they believed in stainless steel, why not use stainless steel for BOTH washers? Or, if they wanted to use plain steel, why not plain steel for BOTH washers?

My friend felt that the threads were too strong to be pulled from expansion due to rust alone. I find it reasonable for both to be contributors -- for chemical corrosion and expansion both working to weaken the threads.

I bet numbers of these folks are few, but I wonder if folks who performs annual or bi-annual coolant changes (despite Honda's marketing of coolant life) have a lower failure rate than those who ran the coolant out to its recommended change interval.

 

Just curious . . .Do you know for sure that the second washer is stainless? When was that determined? Stainless most definitely can add to the galvanic corrosion of the materials around it (particularly aluminum) so if this other (non belleville) washer is indeed stainless then that seems like a poor choice. . .

The belleville washer is an engineered piece with its spring effect being fundamental to its function. The material that it is made from is undoubtedly chosen because it is well suited to being formed into a "spring" washer. Both mild steel and stainless steel belleville washers appear to be available. Aluminum does not appear to be suited to the application. In this case stainless may have been an inappropriate material to achieve the desired "springiness" or perhaps they went with what was cheapest?

One has to wonder . . .Why combine a stainless washer with a mild (not stainless) steel belleville on an aluminum fitting going into an aluminum exchanger?

 

And My Denso is an example of not changing the coolant.

I hadn't noticed until later what the time frame was on changing coolant for the Ridgeline, I expected it to come up on the MM system. But with my low mileage I was 8 years and 64k miles when my first two ATF tests showed coolant in the ATF. Below the 10 years and 120k miles, but after the "initial" change it's 5 years and 60k miles. I'm thinking it should be sooner in the first 10 years.

 

After all the excellent discussion in this thread on radiator/OTW design, I haven't changed my opinion that pull force behind the corroding Bellville washer is just one-of-two fail modes at the root of SMOD episodes.

Based on reports and photo evidence from those who have suffered the tragedy of SMOD, one cause appears linked to the evils of advanced corrosion, the other appears to be mechanical failure of an otherwise healthy looking assembly.

With no conclusive evidence one way or another, using only what remains of deductive reasoning in this old brain - it makes sense to me having one mechanism clamping the OTW to the lower tank and another for clocking the port (see images of OSC & KOYO) is superior to the integrated/single design of the OEM. Is the aftermarket implementation prone to the same corrosion maladies of the OEM? Absolutely. But in terms of pure mechanics, they are superior because they spread mechanical responsibilities across two discrete components, if for no other reason than vibration tolerance spread to two components.

At this point, I don't think ranking these samples is productive or universally accurate. IMO: the Spectra is hands down the most robust among these - but only when it comes to keeping engine coolant and trans fluid from forming the unholy alliance known as SMOD. OTOH, the OTW concentric design and fewer/wider rows in the rad core introduce heretofore unanswered questions. With those unknowns, reduced engine and trans service life is a minor concern.

Setting aside what we think we know about the OEM design, the aftermarket seems to offer two clear alternatives: a marginally improved OTW exchanger mechanical interface employing industry standard heat management - OR - robust fluid isolation with non-standard heat management attributes. A true "pick your poison" scenario.

So, with a fully loaded coarse grained salt shaker close by, here's a new attempt to summarize available info comparing Denso/OSC/Spectra/Koto.

 

With no conclusive evidence one way or another, using only what remains of deductive reasoning in this old brain - it makes sense to me having one mechanism clamping the OTW to the lower tank and another for clocking the port (see images of OSC & KOYO) is superior to the integrated/single design of the OEM.

^^^ Emphasis mine.

This hasn't been discussed to the degree that the other components have been, but I think this is an excellent point. I questioned the use of the thread locking compound on the OEM threads. Is it possible, with the OEM radiator, that the ports are threaded in as far as they will go, and then backed off if necessary to the correct clocked orientation? Maybe they thread it on, and if it's halfway around or more, they continue tightening to clock it correctly, and if it's less than halfway around, they back it off to clock it correctly.

This might explain the use of the thread locking compound (to "glue" the connection in place), and the use of the conical washer -- this would pusha very slightly "loose" fitting away from the tank, pulling the exchanger tight against the tank wall to maintain the larger o-ring seal.

Did the Spectra use a conical washer under the larger nut fastening the exchanger to the tank wall? I can't remember if my TYC did. This is the best picture I have of mine, and it seems to show just one washer.

The conical washer has always intrigued me. Nearly nothing else on the vehicle that is subjected to similar vibrations and stress uses a conical washer. It appears that the aftermarket does not, either. The aftermarket, at least in the Spectra/TYC case, use the threads themselves to maintain the tightness of that connection. But if Honda had to back off the fitting, due to it being a one-piece job and due to it needing to be clocked correctly, that could be the reason for the use of that extra washer.

OhSix, can you examine your OSC's fitting? Does it use two washers or one? Because it doesn't need to be clocked (there's a separate fitting connection), I wager that it may use just one washer.

 

Warning: verbose story telling and linking of personal experience in this post. Read at your own risk. And don't forget the salt shaker.

Hokie, rewinding on this discussion... a minor burr under the saddle has been nagging at my subconscious. Other than normal wear, how does acidity increase in trans fluid (or any other fluid in a closed fluid automotive system)?

Short story:
When I was a kid, a friend of my Grampa owned one of the now defunct "service stations" where people would go to have their cars filled up, regular maintenance performed, flat tires fixed, ETC. His name was Nick - I spent hours there playing with tools and butting in on whatever was happening. He would let me air up tires & top off fluids - while doing that, he taught me the basics of fluid management (like distilled water in batteries and radiators). One of his "tricks" was to siphon the surface of brake fluid reservoirs prior to topping off. When I asked why that was important, he imparted this wisdom:

The air inside the reservoir heats up as temp under the hood rises. That air expands, but eventually cools off. Cooling air in the reservoir pulls in air. Air contains H2O. And airborne particles - which is why brake fluid gets "dirty" over time.

If you don't siphon off the surface of brake fluid and simply pour more fluid in, you are more thoroughly mixing water and dirt, so: topping off brake fluid as pads (shoes at the time) wear = dilution and particle contamination of brake fluid.

Traditional "bleed the brakes" = pushing contaminated fluid closer to pads/shoes.

Contaminated fluid close to pad/shoes = "boiling" when the system is under heavy use & heat propagates from brake assembly up the brake line.

Boiling fluid = greatly reduced hydraulic pressure = spongy pedal - or, if the fluid is hyper-contaminated, can result in total braking failure.

With that in mind - and - omitting the details of where I was and what I was doing at the time (cuz nobody wants to know that :act024, a potential AHA moment entered awareness.

As was taught to me long ago - plain water is the root of most automotive evil. Until you mentioned the conversation with your engineer pal about the anodic index & electrolysis generated from the INSIDE the trans flow circuit, I never connected these particular dots.

It suddenly dawned on me how acidity can change in trans fluid. Evil water.
Just like a *sealed* brake reservoir, engine and transmission fluid systems "breath" to allow internally generated pressures out - which means they can allow air in. In doing that, H2O and particulate matter slowly builds up. The potential controversy around acidic qualities of humid air is deeper than this thread supports, so that needs to be set aside. But to the degree airborne H2O is acidic...

H2O + airborne "dirt" + ferrous materials sheared away from rotating internal components = acidic increase.

Acid in trans fluid = anodic index increase = ever increasing electrolytic properties passing through the system. And that means normal vehicle use is slowly turning up the voltage inducing galvanic corrosion.

So yes, I buy the theory of - in the context of a radiator - increasing acidity in trans fluid movement, inducing internal electrolysis, exacerbated by the presence of three metals, further exacerbated by the external environment and aided by a complete lack of a sacrificial anode is a major contributor to corrosion seen (especially) in the OEM design.

Taking the idea a little farther: most enthusiasts are aware, fluid change schedules are both time and mile related. Even garage and trailer queens spending more time parked than driven need their fluids changed at timed intervals - which might seem counter intuitive. If fluid doesn't age/degrade by using it, why change it out in a vehicle that remains parked for long duration's? Same as above. H2O.

These conclusions may be common knowledge for those more aware than I, but for me, this was a light bulb.

To your earlier point: other than reduction of $ in your wallet, changing fluids at greater than scheduled maintenance can only be beneficial. In light of the increasing anodic index of trans fluid, changing it more frequently may be an additional safety step to mitigate the onset of internally generated electrolytic processes.

I'm talking myself into another layer added to upcoming projects over here. Crap!

 

I think my friend was speaking more to the acidity of the coolant rather than the transmission fluid, but, sure, both fluids will oxidize over time. Oxidation from heat and pressure often results in an increase in a fluid's TAN (Total Acid Number) and a correlating decrease in a fluid's TBN (Total Base Number). In the motor oil world, a used oil analysis often includes the measurement of TBN -- TBN generally indicates the "remaining life" of the fluid and its additives.

http://www.ridgelineownersclub.com/forums/showthread.php?t=116761

The first attachment in that thread shows a report from Blackstone. See the TAN and TBN fields at the bottom -- those are often extra-cost tests and apparently weren't requested in this case. I'm not as familiar with TBN depletion in transmission fluid, nor where DW-1 starts. Most motor oils start in the 8-12 range for TBN, and once the TBN depletes down to 2 or below, the oil is considered "done".

Here's a UOA on engine oil:

http://www.ridgelineownersclub.com/forums/showthread.php?t=98209

See how the TBN is at 3.9...and Blackstone flags anything below 1. That oil is still in great shape, despite the miles.

 

It would be helpful to know the company that supplied the radiators to Honda. I'm sure it is a company that supplies radiators to other car makers.

One item to note is Nissan Pathfinder, Xterra and Frontier owners did win a class action lawsuit about a similar radiator issue.

http://topclassactions.com/lawsuit-...ments/lawsuit-news/2436-nissan-radiator-defect-class-action-settlement-reached/

Unfortunately, I think it is too late and there are too few Ridgeline owners to receive any class action relief for this issue.

 

IIRC, the radiators are listed as Denso.

From that link:
Under the proposed Nissan radiator class action settlement, Nissan will fully reimburse all current and former owners or lessees of a 2005-2010 Nissan Pathfinder, Xterra or Frontier vehicle that paid to fix the radiator or other damages caused by the defect within eight years or 80,000 miles.

Nissan also agreed to pay for future repairs caused by the radiator defect and partially reimburse prior repair costs up to a maximum of 10 years or 100,000 miles. Owners who had the repairs done before nine years or 90,000 miles can receive relief after paying a $2,500 deductible, while reimbursement for repairs done between then and 10 years or 10,000 miles can come after a $3,000 co-payment. This is still a significant savings for Class Members because repair costs caused by the radiator defect can cost $5,000 or more, according to the motion for preliminary approval of the Nissan settlement.

“The use of deductibles effectively caps any class member’s out-of-pocket expense for repairs and relieves the class members of the burden of shopping for the least costly repair,” the motion said. “While the acceptance of the deductibles covers some but not all of the cost of a repair, such a compromise is reasonable since the strength of any class member’s claim is reduced the more trouble-free miles the vehicle has been driven prior to the defect manifesting.”
************************

Ouch. Do I read this right? $2500 deductible. $3000 co-payment? And only 10 years/100k miles?

 

Ouch. Do I read this right? $2500 deductible. $3000 co-payment? And only 10 years/100k miles?

You did. Not such a good deal after all. IMHO - The best remedy is good old fashioned preventative/replacement maintenance.

I know someone who basically sold their Xterra with 110,000 miles for salvage over the issue. The guy hadn't owned the car long before issue occurred. I doubt that he knew about the CA suit. He took a real loss.

 

On Amazon, the OSC 13065 radiator for my 09', they stated.. " Features 4 plate stainless steel coolers that is OEM Standard "

I wonder if the stainless is a typo since many talk here about it being aluminum.
.

 

I bet it's an error in their ad copy. Note, Rock Auto carries the exact same part number (OSC 13065) for $158.

 

I just received a new radiator from Imperial Parts on Amazon, posted a few pics and information here if you wanted to include another brand : http://www.ridgelineownersclub.com/forums/showthread.php?t=53921&page=28

 

Hey OhSix!

If you're still doing Mr. Wizard stuff with radiators, I have an OEM Denso I just took out if you want it. It's in fine shape..... I probably would never have replaced it if I didn't have a fan motor failure (radiator side).
Anyway, if you want it let me know.... I'm here in Carlsbad & can get it to you or you could pick it up.
Nothing wrong with it.... could still be used w/o issue.... or could be lab fodder.

FYI, I replaced w/ new Denso radiator from PartsGeek. I also got "fan assembly" (I chose Dorman) from Partsgeek. I only needed the fan motor, but this assembly comes complete with shroud, fan motor & cable/connector, fan blade.... just pull out & put in... exact match for OEM....and it cost less than just the fan motor from Honda dealer. FYI to anyone interested. Parts got her in about 3 days, install went smooth, save for dealing with hydraulic fittings (line-plugs, hose clamp angles, etc.)... just nuisance slow-you-down complaints.

I too shaved off the foam cushion & used spray-on contact adhesive to reapply to new radiator... worked fine. I did not replace any of the other cushions / bushings (top or bottom).... SoCal is kind to us in that regard.

Anyway... let me know if you want the old radiator.

 

Hey Mr. Nick! Just now saw this post. Yeah man, I'll gladly accept your radiator. Like my Mom's toasters, I love taking chit apart that doesn't need to go back together.

Being a fan of statistics, the larger the sample size the better so let's arrange a hook up and I'll tear that bad boy down for a post mortem inspection. We'll check out arteries, passages and fittings to see how she fared during her service life. Still have a couple cores here ready for recycle. After seeing your post, it occurred to me I haven't taken the top tanks covers off these to see what up. so we can add that to the mix just for fun.

BTW: dang it on me for going OEM on this years fan motor replacement. Didn't even think to look into non-OEM options. Lemme know if you still have the rad - and if so how to proceed. I'm in Carlsbad by Palomar Airport at least once a week.

 

Hey everyone. Thanks for all the great pics and info. My view on all of this is I went with the OEM. I replaced too many things before on other projects in my wife civic that if you don't go OEM there tends to be other side effects. Worse case scenario I'll keep my eye out for corrosion and replace the rad in another 5 years.