Change in steel properties

[tumbleleaves]

Say that a channel is welded to the bottom flange (of a simply supported wide-flange beam): That is, two welds (perpendicular to the beam) across the bottom flange, mid-span where the beam experiences its maximum stress.

Due to the welding and welds (say fillet welds 1/3 the thickness of the flange thickness):
1) Has the yield strength of the beam's flange changed?
2) Has the ultimate strength of the beam's flange changed?
3) Has the fatigue strength of the beam's flange changed?
4) Is this covered in one of Blodgett's books, where?

What I am asking is does welding to the tension flange (or any other tension member) cause a permanent change, that grinding off the weld would not alleviate?

I was wondering about this and really didn't know where to post this question. Asking hypothetical questions at work that aren't directly related to a project is kind of touchy right now--so I'm glad that I can ask it here.

 

[metengr]

Due to the welding and welds (say fillet welds 1/3 the thickness of the flange thickness):
1) Has the yield strength of the beam's flange changed?
2) Has the ultimate strength of the beam's flange changed?
3) Has the fatigue strength of the beam's flange changed?
4) Is this covered in one of Blodgett's books, where?

Replies;

1. Yes, but only confined to the weld region. The weld region consists of the weld deposit and surrounding base metal, aka the base metal heat affected zone.

2. Yes, but only confined to the weld region.

3. Yes, as above.

4. Yes and other welding related references. Use the internet and search.

What I am asking is does welding to the tension flange (or any other tension member) cause a permanent change, that grinding off the weld would not alleviate?

Yes, unless you either remove the surrounding base metal HAZ that was beneath the weld you deposited and ground off or reheat treat the beam. The heat from welding does alter the surrounding base metal.

 

[tumbleleaves]

Based on metengr's response above it does not appear that I asked the questions correctly.

HERE'S MY GUESS OF WHAT WOULD HAPPEN:
I would assume that the base metal would increase in yield strength, decrease (slightly) in ultimate strength; and I'm not all together sure about the fatigue strength (the allowable fatigue stress of steel is the same for different yield strengths of typical structural steel sections?). My guess is allowable fatigue stress would decrease, but perhaps by not a lot?

Again I'm speaking from memory and am not involved in this on a project specific basis, please correct me if I've been misinformed on my assumptions, etc., I will not be offended.

If anyone could direct me to the specific location in one of Blodgett's books where this is talked about? Or some of the specific material processes that are occurring in the metal and to what extent this affects the metal--that would be great. Has anyone ground out the welds in practice and when would this be allowable? What code were you using and where was it addressed in the code? I was hoping to start and interesting thread about a topic I am genuinely interested in.

 

[metengr]

tumbleleaves;

I would assume that the base metal would increase in yield strength, decrease (slightly) in ultimate strength

;

You asked about the change in mechanical properties. The answer is yes the properties will change in the base metal HAZ. In this zone you will get a mixture of coarse and fine grains, a locally hardened region and progressing through the heat affected zone you will see a narrow band of softening. You cannot simply make general statements about the behavior of the base metal heat affected zone from welding because the chemical composition of the base material affects the properties within this zone.

Now, if you progress outside of the heat affected zone, the bulk properties of the steel beam will be unaffected.

I am guessing from your questions that you or someone is contemplating using fillet welds on a tension member of an I-beam. As I stated above, the mechanical and fatigue properties within the base metal heat affected zone will be altered. As to the type and amount of change there is no easy answer other than this – if the heat affected zone is left in this section of the beam, you could potentially have a hardened region that is susceptible to fatigue crack propagation.

To reduce the susceptibility of the remnant heat affected zone after removal of the fillet welds, you either remove the heat affected zone entirely by grinding or locally stress relieve this region to temper the hardened zone (but this will not effect the soft zone that remains within this region).

and I'm not all together sure about the fatigue strength (the allowable fatigue stress of steel is the same for different yield strengths of typical structural steel sections?). My guess is allowable fatigue stress would decrease, but perhaps by not a lot?

The high cycle fatigue strength and ultimate tensile strength are proportional to each other by a factor of 0.5 X UTS. Again, general statements should not be made regarding the performance of remnant heat affected zones. All I can say is that the fatigue properties of the base metal heat affected zone will not be the same as the bulk properties.

 

[tumbleleaves]

Recently this topic was brought to my attention and I was astounded that I was so little educated.

My questions are with regard to structural steel ASTM A36, ASTM A992, etc... welded with E70xx, welded per AWS D1.1, submerged or field welded. There must be some kinds of generalizations that can be made about the reductions in allowable design stresses, generalized or "conservative" as they may be. There must be ways to approach it from specific sections of the design codes. Again is this addressed by Blodgett, where?

No issues with segregation of the material, just a change in grain size? I would not expect the UTS to be linearly related to the fatigue stress, this is counter-intuitive because I would expect brittler with a closer yield to ultimate stress metals to be more likely to crack and thus fatigue?

For 3 years I was an engineer where most of the work involved structural steel typically with welded connections very few bolted connections--that was then. At my present employment I rarely do any sort of design work. The situation that brought this topic to the top of mind was taken care of weeks ago and I was not involved. This is purely a matter of contemplation.

 

[metengr]

There must be some kinds of generalizations that can be made about the reductions in allowable design stresses, generalized or "conservative" as they may be. There must be ways to approach it from specific sections of the design codes.

There are knock-down factors for weld metal in relation to the specified minimum ultimate tensile strength for determining weld size and for fatigue loads in "Design of Weldments" by Blodgett.

See 6.3-6.5

No issues with segregation of the material, just a change in grain size?

Huh?

[/quote]I would not expect the UTS to be linearly related to the fatigue stress, this is counter-intuitive because I would expect brittler with a closer yield to ultimate stress metals to be more likely to crack and thus fatigue?[/quote]

Better read your text books on Mechanics and Materials Engineering.

 

[paddingtongreen]

While the welding is bound to have made some changes to the base metal, there is no requirement to use reduced allowable stress in the beam. If the weld is subsequently ground off, no problem, but if it is burned off, big problem. We were always required to grind out the heat affected metal, sometimes an eighth of an inch and replace it with new weld metal, which, was then ground smooth if needed.

This repair meant propping the beam if it was already in service.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[tumbleleaves]

Interesting!

Based on paddingtongreen's post in this case it was done without any change to the beam design.

I would think there would have to be a check/adjustment (for the allowable stress in the beam) where fatigue was an issue: Say by reducing the allowable stress to whatever the allowable stress of a welded butt-joint would be, since grinding down an 1/8 and filling with weld metal is in a way "a partial" butt joint? This would significantly reduce the allowable fatigue stress of that section of the beam and be conservative?

Or how about this: say the weld metal was just ground off flush and the HAZ left in place then say you neglect 1/8" thickness of the flange since that metal has been heat-affected and analyze the theoretical section at the beam's allowable fatigue stress?...such that one could argue they were complying with code?

 

[desertfox]

Hitumbleleaves

In respect of fatigue, if there are no fatigue plots or data for the beam material, then if you keep the tensile stress below 0.4 * UTS of the material then fatigue should not be an issue anyway.
So if your designing a beam with a safety factor greater than 2.5 fatigue should not occur.

desertfox

 

[tumbleleaves]

SPEAKING OF FATIGUE ONLY:
I think there are two separate factors that I'm contemplating, and trying to separate:
1) When a member is welded to the beam it makes the beam stiffer and causes a fatigue crack at the location.
2) The heat effects and weld metal itself are a discontinuity and change in properties in the otherwise (nearly) homogeneous flange metal and also will affect fatigue.

The various codes (such as AISC) give allowable fatigue stresses when a member is welded on and for various connections (such as a butt). This allowable stress corresponds to the combination of items 1 and 2. What doesn't appear to be addressed is when the weld and member are removed, the allowable stress for item 2 only?--which is my question.

The allowable fatigue stresses given by AISC appear to be much less than 0.4 UTS for when a member is welded to the tension flange? If the member and weld are removed is it then all right to say below 0.4 UTS is good?

Thanks all. I'm enjoying this.

 

[gtaw]

Just a point of information.

A992 is required to be welded using a low hydrogen filler metal / low hydrogen welding process per AWS D1.1.

Not all E70XX shielded metal arc electrodes meet the requirements for low hydrogen. As an example, E7010 utilizes a cellulose based flux covering that introduces an equivelent amount of hydrogen into the weld puddle as an E6010 electrode.

From a practical standpoint only those electrodes that end with 18 or 28 are low hydrogen SMAW electrodes when stored properly. Examples: E7018, E7028, E8018, etc.


Best regards - Al

 

[unclesyd]

Not being well versed in the art and science of beams and things I have several papers on the design of spreader bars using sections other than round and addition of pad eyes to beams. None take a reduction in beam properties for welding other than the normal mandated safety factor for the unit as a whole. The only difference is in the calculation of the safety factor and if you take actual numbers they all workout to a factor of around three.

 

[tumbleleaves]

Thank your for the response unclesyd, greatly appreciate that you took time to write in--and it is interesting that this is the second response where "in practice" fatigue was not considered (or considered negligible?). My guess is that these spreader bars are likely not used enough where fatigue is an issue.

In my experience I also designed numerous hoisting and rigging devices, which was the most challenging design work I did. I worked with and reviewed the work of engineers, some that designed spreader bars throughout their whole career and are now retired. Many of these engineers never considered fatigue, and I did not see calculations by others where they magnified the moments either (this was typically negligible (about a 5% or 10% increase in stress typical).

My understanding of my work at the time and of the design guidelines and codes was that the design of "below the hook" devices complies with AISC and a complete analysis will treat fatigue stresses in the same way as specified in the AISC manual.

I agree safety factor typically ranged from 2 to 4, depending on the type of design.