Some questions regarding TIG welding an aluminium inlet manifold

  1. melle Member

    Messages:
    348
    Location:
    Bath UK/ Dublin IRL
    I have a plan. Could do with a bit of feedback though to make sure it doesn't end in tears. ;)

    The plan is to shorten a Ford 2.9 V6 EFI intake manifold and plenum so they fit a Cologne V4 engine for one of my Saabs. I've done my homework, all relevant ports and channels line up and I'm convinced I can make it fit. I need to hack a 12 cm piece out of the plenum and the manifold (that's how much the V6 is longer than the V4), and then the remaining pieces need to be welded back together. The cooling channels need to be modified too. I'm set up for MIG welding aluminium, but I've never had much luck with it despite having a decent welder, gas and wire, so it must be lack of skill. I don't have a TIG welder myself, nor do I have any experience with TIG welding aluminium, but fortunately a mate is happy to do it for me. I've read a few threads on this forum on welding cast aluminium manifolds, but I still have two questions.

    Comparison between a V6 manifold and a V4 one, it's clear it isn't a bolt on conversion!

    [​IMG]


    So this is the idea, V6 manifold/ plenum in the first pic, rough Photoshops of the envisioned end result in the following two.

    [​IMG]
    [​IMG]
    [​IMG]


    After cutting out the middle, I can bolt the pieces to a jig (e.g. a thick piece of flat material, or maybe safer, an engine block) and align them, and then tack them with the MIG so my mate can fully weld the manifold in his workshop. My first question is, how much warping should I expect after TIG welding the halves together? I will have the contact surfaces machined after welding, but there's not that much leeway because the manifold can't sit too low between the heads.

    Then onto the cooling situation; the V4 has a coolant channel under the inlet that the V6 doesn't have. Fortunately there's empty space there and blind ports have been cast in and machined, so it looks like I can modify it to do what I want it to.

    [​IMG]

    Before cutting out the middle of the manifold, I'll remove the remaining cross web with a die grinder, and mill down the ones that run along the length a few mm so they're flat. I don't have a mill, but a jig and a router has worked very well for me before on aluminium (decent router bit or carbide burr and plenty WD40 as lubricant). The idea is then that a "lid" (approx. 60mm wide) can be welded on when the two halves have been welded together, as per the following pic (the piece that isn't filled in will go).

    [​IMG]

    My second question is what grade aluminium I should use for the infill piece. It probably doesn't matter much for strength, but it might do for bonding to the cast aluminium? Can't do with cracking welds as it need to be fluid tight. Not sure if corrosion resistance matters because there will be coolant flowing through it? I think I'll just look for a short piece of flat stock on eBay, 3-4mm max thickness should be OK I'd think? Maybe I'm overthinking this (or quite the opposite!), I just completely lack experience in this domain. Thanks in advance for your feedback!
     
  2. Bobtail4x4 Forum Supporter

    Messages:
    817
    yorkshire
    do you NEED efi on a V4?
    I would chop the top off the efi fannymold and graft it on a V4 one,

    you do know you can buy a efi kit to fit on the single webber inlet?
     
  3. minimutly Member

    Messages:
    1,512
    Location:
    Pembrokeshire Wales
    You need all the help you can get on that v4..,
    On a sensible note - your trying to do something very difficult, which will likely end up in the bin. You will need to weld old dirty aluminium, and you really need to have clean metal both sides really.
     
  4. grim_d

    grim_d Unlikeable idiot.

    Messages:
    3,104
    Location:
    Scotland - Ayrshire
    This needs Allen Millyard's brain.
     
    Hopefuldave and tom2207 like this.
  5. tom2207 Member

    Messages:
    2,522
    Location:
    uk northern ireland
    he would just do it with a hack saw the guy is inspirational
     
    Hopefuldave likes this.
  6. Brad93

    Brad93 M J B Engineering

    Messages:
    10,404
    Location:
    Essex
    I’d be worried about distortion. Particularly if you are lacking in amperage.
     
  7. melle Member

    Messages:
    348
    Location:
    Bath UK/ Dublin IRL
    No, but I WANT EFI on a V4.

    That was my first thought, but it's easier said than done.

    I know, but its 1) bloody expensive, 2) no fun at all and 3) I want an injector per cylinder.

    It's a Cologne V4, not an Essex V4. ;)

    That's very well possible, but then at least I've had some fun trying to make it work!

    I'm not the first one trying to weld a manifold, so we'll see.

    Lack of amperage is not an issue, but distortion is something that worries me too as I indicated in my opening post. Would it help to pre-heat the parts?
     
    Parm likes this.
  8. Brad93

    Brad93 M J B Engineering

    Messages:
    10,404
    Location:
    Essex
    No preheat doesn’t help at all.
    I’d say 300A to weld that properly with minimal distortion. Ive also been informed by someone who does this sort of modification day in day out, that without mixing a bit of helium if you’re fighting a losing battle with distortion.
     
    Richard. and minimutly like this.
  9. melle Member

    Messages:
    348
    Location:
    Bath UK/ Dublin IRL
    My mate is a very experienced TIG welder and he has an impressive industrial type set-up which should have plenty oompf, but he normally only welds clean new ali. He's happy to give it a shot, but maybe I should find someone who's a bit more experienced with this sort of stuff.
     
  10. minimutly Member

    Messages:
    1,512
    Location:
    Pembrokeshire Wales
    Reallistically, if you want to do this, you'll have to cut one up to see if its doable. You'll need to mate them where there isn't a void. I'm not sure if you would butt them, butt them with Vees, or leave a gap to fill - but I would go with option 1 I think.
    Bolt in place with sheet steel shims instead of gaskets, tack everwhere possible, remove and weld. If it distorts you'll have to have it machined, which is no biggie.
    No mean feat, but it's not a kawasaki 6 pot!
     
    Hopefuldave likes this.
  11. Brad93

    Brad93 M J B Engineering

    Messages:
    10,404
    Location:
    Essex
    Give it a try by all means but personally I’ve done one manifold and one sump, and won’t do anymore. Not without 300A and helium.

    I tried a 20mm thick jig and 100 degrees of preheat. Had 3mm of bow in the sump.

    The head/manifold was bent in all sorts of places but luckily that could be milled off.

    You want the best fitup possible with no gaps. Any gaps promote distortion.

    the sump had massive gaps because it was fabricated and the parts stuck together with tape by the customer.
     
  12. melle Member

    Messages:
    348
    Location:
    Bath UK/ Dublin IRL
    That's exactly the plan! It's just a fun project and if all fails not much is lost.

    @Brad93, if it would bend 3mm there's no way I can get it machined so it still fits between the heads I'm afraid, so it's worth doing some more research on how distortion might be avoided. Thanks for your input so far!
     
  13. Brad93

    Brad93 M J B Engineering

    Messages:
    10,404
    Location:
    Essex
    I have no pictures of the manifold. But do have some of the sump.

    you can see the steel plate it’s bolted down to.

    doesn’t help when the material is dirty which most used cast car parts inevitably are.
     
    • BA44F6D5-2BC5-4256-8F13-CEC3541B65F5.jpeg
  14. minimutly Member

    Messages:
    1,512
    Location:
    Pembrokeshire Wales
    Nice welding there! But I would make a point - there are quite a few welds passing perpendicular to the length of that sump, each one shrinking as it cools, contributing to the bowing. With your manifold there will only be one weld.
    On this point, one thing I allways do if welding a patch or plate - I never tack all four sides, but weld fully one side, let it cool, then move onto the others - that way you only have shrinkage from one joint.
     
  15. Wallace

    Wallace Member

    Messages:
    7,529
    Location:
    Staines, Middlesex, England.
    Have you considered copying the angle and position of the V6 fuel injector and welding bungs into the V4 manifold to accept fuel injectors then adding a converted 38DGAS as a throttle body? Alternatively, as opposed to cutting the centre out and welding the two halves, look into cutting the rear two cylinders off the V6 and building up the valley gasket then reroute the cooling system to replicate the V6 as the heated manifold is for the carb benefit and not necessary with EFI.
     
  16. Brad93

    Brad93 M J B Engineering

    Messages:
    10,404
    Location:
    Essex
    Was unavoidable in the thickness that was supplied.

    Whether you tack and let cool or weld it and let cool before welding again the heat input and contraction of weld metal will be the same.

    I jumped from area to area and utilised a back step method.

    If i had welded one side fully all the tacks would of cracked opposite. Without using 1” long stitches, which are then hard to blend in when you weld over them.

    There is only two sure fire ways to reduce distortion.

    • Reduce your heat input, by making sure the fitup is good and your travel speed is high. This is where the helium mix comes in
    • Preset the part. I never would of expected the sump to bow by 3mm. I probably should of put a 2pence piece under each long edge and clamped it down. Which would be 1.5mm or so. A 3mm packer would be seriously hard to pull the part down and would then form more gaps.
     
  17. melle Member

    Messages:
    348
    Location:
    Bath UK/ Dublin IRL
    I'm also building a V4 K-Jetronic (Bosch mechanical injection) engine, and that's the approach I'm taking on that one. I'm using a Porsche throttle body, which happens to fit the stud pattern of the original carb. I know someone who's just drilled holes in the manifold with a cordless drill and epoxied the injectors in, and that works absolutely fine, his car is a beast to drive. I just want to see if I can make this work, not because it's the simplest or best solution.

    I have considered that and I may look into it again, but I think it is way more complicated than what I currently plan to do.

    That's a good point and re-routing the cooling around the manifold is something I considered. I have a diagram, but I'm still not absolutely sure I need the coolant flow from back to front, or if the V4 manifold is just designed that way to warm it up to benefit fuel atomisation. If this part of the cooling system is necessary, I think having the coolant flow through the manifold is the most logical option, but I'll definitely give it some more thought too.
     
  18. minimutly Member

    Messages:
    1,512
    Location:
    Pembrokeshire Wales
    Note for @Brad93 - I won't pretend to approach your welding ability, so hats off for that. But if I was going to make that sump I would have tacked it all up, then welded the perpendiculars, then removed the whole lot and if necessary pulled it back into shape, refitted, welded one end fully, then worked down allong it, leaving the final perpendicular weld to last.
    It doesn't quite fit in with your one of two ways to reduce distortion (maybe does the second one?), but it would have worked. Oh, and you're wrong - welding one side (or rather end), letting it cool and then welding the other end does reduce contraction - by exactly half.

    Note for @melle - efi doesn't need a hot water jacket, so if Ford kept them, was it a historic thing, or does the engine need them?
     
  19. melle Member

    Messages:
    348
    Location:
    Bath UK/ Dublin IRL
    I know a heated intake manifold is not needed for an EFI engine, but as I understand it the coolant from the rear of the V4 engine normally flows through the manifold to the thermostat housing; I can't just cut that I'm afraid. Ford didn't keep it, that's why I'm trying to recreate it, see photos in opening post.
     
  20. minimutly Member

    Messages:
    1,512
    Location:
    Pembrokeshire Wales
    I don't see a stat mounting on the efi manifold? (Obviously you don't need one, you could fit one in the top hose), but you'll still need a path out for the coolant? I see that making a path in the manifold would work, but still you need to get it out?
     
Advertisements