Limit for Throat Thickness of fillet weld, Inspectors View? 5

[IJR]

My dear friends

At design time, we limit throat thickness to 0.707x min plate thickness, which implies that length of the leg can not be taken to be more than the plate thickness

But that is for calculations.

I am looking for a perspective from Welding Inspector's view

What is the concern if leg length happens to be somehow larger than the thickness of the plate(in other words throat thickness is larger than the above limit)

Allow me to post this one too, to "structural engineers other issue"

Respects.
IJR

 

[HgTX]

There is some discussion in the crosspost at thread507-148680

Hg

Eng-Tips policies: faq731-376

 

[gtaw]

Few welding standards address fillet welds sizes larger than those specified by the welding symbol. One that I did encounter was in a military ordenance standard and it was only for fillet sizes less than 3/16 inch (that's based on memory alone). Another one that might have a limitation is AWS D17.1 for aerospace applications and again, I believe it is for welds that were rather small. Most welding standards are concerned with the fillet weld being smaller than specified.

Excessively large fillet welds cost more due to the volume of weld metal deposited. Consideration must be given to the additional distortion introduced. Distortion caused by multiple pass fillet welds will typically be greater than single pass fillets of the same size due to the increased number of thermal cycles of multiple weld passes.

AWS structural codes (D1.1, D1.2, D1.6, etc.)limits the size of the fillet weld along the edge of a plate if the plate is more than 1/4 inch thick. The maximum fillet weld size is 1/16 inch less than the plate thickness if the plate is over 1/4 inch thick. I have been told that this was for inspection purposes. The standard fillet weld gage is difficult to use, i.e., obtain an acurate reading, if the weld is the same as the thickness of the plate. Assuming the weld is specified as the same size as the plate thickness, the welder could melt the edge of the plate so that the actual weld is something less than the size specified. Without a defined plate edge, there is no reference point to use to measure the weld size.

Best regards - Al

 

[unclesyd]

Sometimes a fillet weld is made with deliberately long legs and is specifically call out on the drawing with by the symbol or the notes.

We call out a 3 to 1 fillet on quite a few fabrications to mitigate fatigue problems. Most of the time it is required to be ground after welding.

One of the worst problems on any fillet weld is when you have a cold lap on the leg of any size weld. I have probably seen more failures in dynamic systems from this than all other defects combined.

 

[IJR]

unclesyd

What is a "cold lap"?

Trying to get the best from your post-since you seem to be working on fatigue related designs.

Why long legs?

respects
ijr

 

[unclesyd]

In respect to fillet welds the Cold Lap is where the weld metal doesn't melt/fuse to the base metal and just rolls over it. The cold lap is primarily but not limited to the thicker metal side when using equal leg fillet welds between two different thicknesses of material.
This normally leaves an extremely sharp notch at the outside edge of the fusion zone under the lap.

I always call out the elliptical fillet when there is concern that there is a possibility or has been a case of fatigue failure. This especially successful in the thick and thin case or the big and little.

This works extremely well in cases like fan and blower blades welded to the hub, stub shafts, some pipe connections.

It takes a little practice on the welder's part to get use to slightly offsetting the weld profile.

I try to achieve a 3::1 eclipse with both toes blended smooth. The 3 being on the thin or little part.

 

[gtaw]

Cold lap is also called overlap (AWS term) and roll over by welders.

Al

 

[GenB]

IJR

the 0.707 x min plate thickness, is the minimum required,
and is a little more complicated thatn that because you have to proove that further reinforcing is not required, and if it does, a larger weld may be needed.
genb

 

[IJR]

GenB

Are you sure 0.707xtmin is the MIN?. I am not sure about that, because that is simply saying that 70.7% of the thickness is the minimum for fillet.

As far as I know the minimum is 3mm.

respects
IJR

 

[weldtek]

.707 of the leg length is the effective throat, for any size fillet.

 

[GenB]

IJR, it is the minimum without reinforcing, if you need reinforcing you may need a larger fillet.
you have to calc/proove your welds to see if a bigger fillet is needed.
at this time I do not know of a maximum fillert but we use fillets quite large for even thin metals;
1/4; 3/8 or even 1/2in.
genb

 

[IJR]

GenB

You seem to be working in a specialty. So I am going to bother you a bit more

The max 70.7% of thickness for welds is actually a design limit. But as pals have suggeted in this forum, roots do become larger in practice, and my post was about how much larger is acceptable. European standards put a limit of 3 mm in excess of the root, and the root is not to exceed 15% of the design value. But these are design related considerations.

Please go on tell us of a code or literature related to reinforcements

respects
ijr

 

[henri2]

IJR, you are right about the 3 mm minimum fillet weld size for statically loaded structures. It is indicated in AWS D1.1 Table 5.8 Minimum Fillet Weld Sizes.

Now to your original question about the inspector's view point on the subject you raised. All the inspector can do is perform inspection according to approved plans, specs and applicable codes..and contact the engineer for clarrification. Some codes do not have a plus tolerance for the size of the fillet weld; therefore this is information that IMO should be included in the job specs and or plans to minimize distortion etc. Even if there is no plus tolerance for fillet weld size, a well trained and experienced welding inspector may raise some concerns if the weld sizes are too excessive because he/she will understand the consequences. Regarding undersize fillet welds (underun), D1.1 for instance has provisions in Table 6.1

weldtek you write ".707 of the leg length is the effective throat, for any size fillet". For the typical non-tubular connection (that excludes situations like skewed T-joint) where members are joined by flat or convex fillet welds this is essentially a true statement because the leg = weld size. However, if the weld is a concave fillet weld, the weld size and actual leg size will not be the same and that is why it is important for the welding inspector to use the proper fillet gage to determine the size of the fillet weld.

 

[weldtek]

henri2,
You are absolutely right, I appreciate the correction, and would add that Annex I and II of AWS D1.1 provide further clarification.

 

[SERJI]

Now I can´t see the standard, but in standard ISO 5817 you have the answer (guidance only because)

 

[rogerheadford]

Dear Forum,
AWS allows the throat for a partial pen. fillet weld to be the shortest distance from a point connecting the ends of the legs.
Can this also be considered for a simple fillet weld ?.
In addition, if the profile of the weld is convex, can this additional weld be considered in evaluating the throat?? (i.e. 0.707 x leg + convex)

 

[JR97]

I want to amplify what GTAW wrote.

For discussion assume the fillet weld is to join two plates laying on top of one another with the top plate offset - the callout is for a continuous fillet weld of size equal to the top plate thickness.

As GTAW stated, if the fillet is the same size as the top plate it is impossible to measure the size of the leg on the bottom plate. The welder will inevitably melt away the top corner of the top plate. Without knowing where that corner is, there is no way to determine if there is sufficient leg size on the bottom plate.

In our shop before welding we will put reference marks on one of the plates 1.0 inch from the corner. After welding, the inspector can readily determine the size of the leg on the bottom plate by starting his measurement 1.0 inch from the reference marks.

 

[gtaw]

"JR97" makes an excellent point. His solution to the problem of "edge-melt", as it is called by MIL-STD-2035, uses the same approach as required by the Navy when welding socket fittings.

Best regards - Al

 

[JR97]

rogerheadford,

In response to your question about convex fillet welds -

Using the definitions of AWS A3.0, the distance from the bottom of the root to the face of the weld is the actual throat. The effective throat is the same distance excluding the reinforcement.

Remember the fillet weld size is the leg length of the largest right triangle that can be inscribed within the fillet weld cross section (again from AWS A3.0). Surface convexity does not increase fillet weld size. From a strength standpoint the convexity does not add to the strength of the weld either.

An inspector will measure the size of a fillet weld by first determining the shape of the weld (convex or concave). If the weld is convex, the inspector will measure the leg lengths with the smaller length being the fillet size. If the weld is concave, the inspector will measure the throat with the fillet size being 1.4 times the measured throat.

Of course different formulas apply for uneven or skewed fillets.