Over Molded Inserts - Coolant Manifold

[AFish66]

Hi all,

I apologize in advance if this is not the correct thread category. I have an application for a coolant manifold (50/50 ethylene glycol/DI water) that will experience max temp of around 70C and max pressure around 14psi and am in need to create this out of a thermoplastic material. To do this successfully, I also need to incorporate two locations for an NPT thread pressure and temperature sensor (1/4" NPT and 1/8" NPT respectively). I believe I am limited to these sensors at this time, but was wondering if there was much experience with overmolded NPT inserts. They do not seem to be readily available in that thread style in the market, so I am guessing it's not very common. I would assume there are issues with overtorque/leaking, but do not think that will be that big of a problem since this is low volume and we have skilled techs building our assemblies. Any experience on using an overmolded NPT insert and where can I find them? The thermoplastic material will be PVDF.

Thanks

 

[Pud]

You will probably have get inserts custom made, or mould threads into the part.
Q: why are you using PVDF when there are hydrolysis and glycol resistant nylon 6.6 grades available as used by almost, if not all car manufacturers?


Politicians like to panic, they need activity. It is their substitute for achievement.

 

[FACS]

Yes,
I have experience doing this with PVC injection molding and a 1.5-inch NPT stainless steel insert.

The two issue in this case were the torque applied during installation id NPT fittings and sealing the insert from leaking.

The torque issue was not easy to solve because this was a field installation into water wells. The PVC was a diverter spool and the plumber would crank on the part with a huge pipe wrench. Endless FEA and rib designs were tested before we were able to establish a design that had a reasonable torque value.

I don't expect 1/4 and 1/8-inch NPT will give you that much grief, but you will have to do a lot of analysis on leaving a thick boss.

The leaking is the hardest issue. Putting a insert in this will require you to work with the molding company. The part may have to be heated before molding and it will have to be designed to allow material into areas to stop leaking. Bubbles and poor plastic conditions will cause leaking under pressure. Also the expanding of the metal insert due to the high heat will be greater than the plastic, that will be a problem.

In my case, I had to design the mold with a special insert that allowed the plastic to mold over the top of the insert and be part of the thread. So that when the part was finished, the fist three threads of the boss were plastic, and then that transformed into stainless. We did this because the transition or knit line of the plastic holding in the insert was never good enough to stop leaking. Including the plastic threads to join the metal threads allowed us to use the liquid tight sealing to cover the knit line.

I can try to get you a picture to show you what I mean.

 

[Artisi]

Maybe I'm being too simplistic but look at PVC pipe fittings with threaded brass inserts, seems it can be done quite easily.
These fittings seem to tolerate heavy handed plumbers screwing fittings into them.
Talk to the PVC pipe manufacturers for ideas / assistance.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[AFish66]

Hi all,

Thanks very much for all the replies. Nylon would be my first choice, however the PVDF material is 1 of 2 compatible thermoplastics listed as approved for this specific application. I would attempt to mold the female NPT threads, but since we are threading in a metal MNPT sensor, I am worried that I may have more problems with cracking the plastic - maybe this is less of a risk than the risk of leaking with using an insert?

@ FACS, thank you very much for the information, the picture makes sense. Perhaps I can just get the mold shop to create a design like this. I will see what they say.

@ Artisi I have seen some "off-the-shelf" NPT inserts that are overmolded with PVC, and I was thinking it would be easily done at first. However, what made me second guess this was looking at some old automotive thermostat housing designs with overmold inserts, and then realizing the design was changed later to utilize an o-ring seal for the sensor location. There were two instances I found of Ford parts like this, which made me assume they had some leaking issues.

 

[FACS]

With that said, the issues are always with NPT fittings as they are the most often over tightened fittings. Once someone cranks on then the bond between the plastic and insert are compromised. Thus the eventual leaks.

You will have to be sure to pay attention to Thread design for NPT as the effective thread length needs to be covered. The insert and the plug to hold it during molding will all be custom designs. The insert cannot have the full effective thread on it.

Good luck

 

[FACS]

"Maybe I'm being too simplistic but look at PVC pipe fittings with threaded brass inserts, seems it can be done quite easily." _ yes, too simplistic of an observation unless you are out in the field installing these fittings. Unless you have designed these devices, unless you have real life experience that helps this post.

"These fittings seem to tolerate heavy handed plumbers screwing fittings into them." _ No, they do not. You are making an incorrect assumption. I had pages of data that proved over tightening is common occurrence for some people and NPT threads. I went into the field and watched them. It only takes one leaking fitting buried under 12-feet of earth to curse people that make these assumptions without any data or experience.

"Talk to the PVC pipe manufacturers for ideas / assistance." No, talk to your mold maker and plastic injection molding firm. They will assist you in the right design.

If you haven't done it, don't offer advice.

 

[EdStainless]

I used to have some parts molded in a higher strength resin (40% glass PPS) and there were some threaded bosses.
To reinforce these we molded a SS collar around the boss. This let us make the threads in the plastic, and we got good sealing, but the SS was there to back it up and limit the tensile stress on the plastic.
We cross drilled a hole in each sleeve so that they would lock in place, we just used thin SS tube (0.035" or 0.049" wall).
Since thermoplastics have larger CTE than metals the collar gets tighter when the parts heat up.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy

 

[FACS]

EdStainless,
That is actually a better idea for smaller threads such as this. AFish66 should consider that.

I did try that for the PVC project I described above, but the PVC still cracked on large NPT threads. Maybe it wasn't thick enough. Glass filled PVC also broke. I don't think that would happen with different plastics and small threads. Nylon is going to handle this better as it is more forgiving than PVC, depending on the nylon type.

 

[AFish66]

I would definitely be willing to try it. I will draw up a concept and share it to make sure I'm understanding correctly

 

[pylfrm]

AFish66,

Regarding the metal collar idea, this Spears product brochure may be of interest. They claim a patent.


pylfrm

 

[Artisi]

FACS
Thank you for your expert advice:

If you haven't done it, don't offer advice - - wrong.
You went into the field and watched, that's great - you are now highly experienced.

Who are the PVC pips manufacturer's other than the injection moulding/ers and die manufacturers?

Did you actually read the OP, 1/8 and 1/4 inch fittings, a rocket science requirement.

Something like 90% of Asia have PVC plumbing, the majority installed by marginally skilled personnel - sure, some failures but nothing is perfect.

My suggestion is when people spend their time even with "a simplistic" idea for the thought process, you should should accept it for what it is.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[AFish66]

@ FACS and EdStainless,

Thanks for the continued support. Here are some images of the collar insert concept. Please let me know if this is the intent.

 

[AFish66]

Here's a second image.

 

[EdStainless]

We didn't capture the bottom edge, in fact we left a large radius in the plastic there.
The metal sleeve shouldered against a step in the mold up about 1/4".
As I recall we tried both smooth tube and grit blasted finish, it didn't matter since the function does not rely on adhesion.
The sleeves did have to be warmed before going into the tooling. The die is made to the 'hot' size. Otherwise they will expand as they heat up and you can't get them out of the mold.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy

 

[AFish66]

Thanks Edstainless. Check this out, I just found a solenoid valve that uses FNPT fittings with the stainless collar like you mentioned.

https://www.burkert.com/en/Media/pl...-EU-EN.pdf?id=DTS0000000000000001000011072ENK

 

[FACS]

It looks to me like you two got this ..
Ship it.

 

[EdStainless]

We never thought about a patent back in '92 when we did it, assumed that it was common sense.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy