Welding Tips and Tricks

Hi all, I'm looking to add an oxygen sensor to a 1970s cast iron exhaust manifold and would like to build up some thickness for machining. Considering thermal expansion differences, would it be prudent to use either silicon bronze or aluminum bronze for tig brazing?

Thanks in advance.

Chris

Save yourself a headache and put the sensor in the pipe.

sedanman wrote:Save yourself a headache and put the sensor in the pipe.

Amen to the above.

Additionally, a 37 year old +, CI manifold just may not accept any form of welding, depending
on how much oxidation and 'cooking' it's already seen--which 'isn't any big deal' if one's never dealt
with grainy, degraded CI manifolds before.

I was rather hoping to install the O2 sensor in the manifold to put it up stream of any leaks in the coupler, which it is very prone to.

I was thinking maybe since it's just buildup and not trying to hold the cast together, that it would make it more doable than trying to repair a cast manifold.

Thanks for the input guys, I'll definitely take it under advisement. Anybody else?

chrisharon wrote:I was rather hoping to install the O2 sensor in the manifold to put it up stream of any leaks in the coupler, which it is very prone to.

I was thinking maybe since it's just buildup and not trying to hold the cast together, that it would make it more doable than trying to repair a cast manifold.

Thanks for the input guys, I'll definitely take it under advisement. Anybody else?

You should have the O2 sensor within 12-16" of the cylinder head as exhaust gas temperature factors greatly into the sensors ability to read. Factory fuel injected cars that have pre and post cat O2 sensors have them directly in the manifolds on newer stuff, but they are also designed with that in mind. I'd imagine you are retrofitting fuel injection onto an old engine. I understand your point about wanting to eliminate false lean conditions because of the leak, but that is typically going to be a less than ideal location in my opinion.

Sent from my VS995 using Tapatalk

GreinTime wrote:

chrisharon wrote:I was rather hoping to install the O2 sensor in the manifold to put it up stream of any leaks in the coupler, which it is very prone to.

I was thinking maybe since it's just buildup and not trying to hold the cast together, that it would make it more doable than trying to repair a cast manifold.

Thanks for the input guys, I'll definitely take it under advisement. Anybody else?

You should have the O2 sensor within 12-16" of the cylinder head as exhaust gas temperature factors greatly into the sensors ability to read. Factory fuel injected cars that have pre and post cat O2 sensors have them directly in the manifolds on newer stuff, but they are also designed with that in mind. I'd imagine you are retrofitting fuel injection onto an old engine. I understand your point about wanting to eliminate false lean conditions because of the leak, but that is typically going to be a less than ideal location in my opinion.

Sent from my VS995 using Tapatalk

Also, leaks directly downstream can cause false lean conditions as well because there is reverbrating pressure pulses that can draw atmosphere in going the other way. It's a lose lose either way with an exhaust leak within a foot or so of an O2 sensor

Sent from my VS995 using Tapatalk

chrisharon wrote:Hi all, I'm looking to add an oxygen sensor to a 1970s cast iron exhaust manifold and would like to build up some thickness for machining. Considering thermal expansion differences, would it be prudent to use either silicon bronze or aluminum bronze for tig brazing?

Thanks in advance.

Chris

...so--rather than use a machined SS coupling to thread the O2 sensor into-then have coupling welded to the manifold (with coupling removed for the welding--suggest)....your plan is to build and then drill/tap the buildup and thru the CI layer.

You need to try SiBrz build up, drill/ tap on a scrap manifold first. The puddle formed--and the CI substrate,
even with relatively little filler dilution can ramp up into the Rc 60's hardness range---pronto; along with the heat
forming ungodly carbides, etc.--which makes things really not too machinable.
99 nickel, 312 SS, etc. fillers can and will show this dilution problem as well.

You're hell bent to make this as difficult a job as possible.

Other than differing expansion rates, I have no idea about the metalurgy, and hadn't even thought about it. Thanks for pointing it out.

Once upon a time, some 25 years ago, I knew there was a 'sweet spot' for sensor placement, but only remembered having seen 80s vehicles with the sensor located at the collector (?) portion of the manifold.

OK, thanks to ALL who replied, I do believe I've been convinced. I'll leave the manifold alone and put a sensor bung in the exhaust tubing.

Thanks again to those who replied.

Chris

chrisharon wrote:but only remembered having seen 80s vehicles with the sensor located at the collector (?) portion of the manifold.

Yup.. You are correct.. They had to at the time as many sensors in the late 70's and early 80's did not have heating elements built-in and relied purely on the exhaust gas heat to get to the operating temperature. (exhaust oxygen sensors don't work unless they are heated to over a critical temperature)

So to get them hot enough within a few minutes they had to get very close into the exhaust manifold and near the ports.

Current sensors are all heated with an electric element so don't need this close proximity for that reason anymore (they may do for other reasons like packaging or emission requirements on speed of response)

Even worse.. Wideband sensors that are now common for new builds and retrofits actually have a definite upper temperature limit for the sensor body to operate in and survive, so these often need to be fitted a little more downstream unless extra heat insulation is fitted to combat this (OEM installs do that with 'encased' manifolds and the like).

LIke mentioned above.. I suspect your best plan of action would be to fit a sensor in a downpipe after the manifold and if you are worried about leaks on flanges (which can definitely happen and mess with your AFR measurements), then focus your efforts on that to make sure they are in good shape. Eg. make sure all mating surfaces are clean, not pitted and flat (belt sander works wonders!) and use a good gasket (multi layer steel if available) and good fasteners.

Good luck!

Bye, Arno.

This may be a little late to the party but I would preheat the entire manifold to 4-500* and tig 99% nickel stick rod after removing all the flux coating and buffing down the cast to a clean and shiny surface. I would then wrap the manifold in fire/heat resistant material, an old leather welding jacket works great. Leave it be for a couple hours and it should be good to go. It's a little unconventional but it's a tried and true method I've used many times for welding cast manifolds.

I've decided to move the O2 sensor down stream, maybe as far as where it Ys together.

I do have an 'ear' that could use some building up though, so thanks for your post.