alberto

Minimum and maximum fillet weld thickness.

27 posts in this topic

Hi!

What is the minimum and maximum thickness a fillet weld should have?

Thanks in advance!

Share this post


Link to post
Share on other sites

According to EN 1993-1-8 section 4.5.2, "The effective throat thickness of a fillet weld should not be less than 3 mm".

As far as the maximum throat thickness is concerned, eurocodes do not mention anything!

However, a very good approach is: maxa=0.7*min plate thickness.

Share this post


Link to post
Share on other sites

I'll chip in my two penny worth as well...

The thickness of welds, especially fillet welds, may be controlled by things other than the code. If you have a very thick piece to weld, it may 'quench' the weld and lead to brittleness, even with pre-heat. I would never, and this is just 'personal taste', coloured by many 'near misses' over the years, use a weld less than 5mm in normal constructional steelwork. There are always exceptions though.

I have seen failure of a 6mm fillet weld during erection when the HSFG's where torqed up - the cause was a combination of welding distortion and small welds compared to the material size and thickness. The solution to this was to grind out the welds and replace them with 15mm fillet welds. This was way way over the strength required but the intention was to put in plenty of heat. 15mm site welds were no easy matter as the welder, working on a platform we had to erect with shelter, had to run transverse to the weld, back and forth, rather than long continuous longitudinal welds runs. As an aside, I'll mention that we set fire to the roof of the building we had built below. Ah, how we laughed...

As for maximum size, I feel that it's dictated by circumstances with no maximum. If you have really big welds, I think there is a possibility that there may be a better way of doing it!

Share this post


Link to post
Share on other sites

For maximum you take 0.7 of the plate if the plate is welded only from one side and 0.5 if it is welded from the two sides, for it is considered that with more than this you will just burn the material and you will not have a real result.

For minimum sqrt(t.p.max)-0.5mm.

I do not remember, but I think this information was taken from DIN.

Greetings

Share this post


Link to post
Share on other sites
I would never, and this is just 'personal taste', coloured by many 'near misses' over the years, use a weld less than 5mm in normal constructional steelwork.

Given the fact that the maximum throat thickness is 0.7 of the plate thickness, either it means that you don't use plates thinner than 7mm or it means you surpass the maximum throat thickness for plates <7mm... Although I agree with being conservative over the welds, I usually use the maximum value 0.7t as a general rule (in Romania the men-hours are still cheap).

As for maximum size, I feel that it's dictated by circumstances with no maximum. If you have really big welds, I think there is a possibility that there may be a better way of doing it!

You have to take in consideration that with a>0.7t, the heat generated will create a bigger Heat Affected Zone (HAZ) which will affect a relevant portion of the thickness of the plate. That area is weakened and may be the cause for welding failure (although is actually plate failure).

weld_decay_weldment_corrosion.gif

Share this post


Link to post
Share on other sites

In these environs, I try to use 7/16" or approx 11mm weld as a maximum if I can. It's about the largest fillet weld that can be done in a single pass. The standards here also stipulate a minimum weld size based on the material thickness. I'll see if I can scan the relevant pages later.

For HSS sections, CIDECT research indicates a max fillet weld size equal to approx 10% greater than the tube thickness. Testing has shown this approximates a nearly full penetration weld.

Dik

Share this post


Link to post
Share on other sites

I need help with welded RHS sections. By CIDECT design guide 3 (p.3.9) minimum throat thickness is

a>=0.92t for S235

a>=0.96t for S275

a>=1.1t for S355

These numbers are a lot bigger then amax=0.7t that is mentioned here before. And I do not understand why this CIDEST design guide 3 does not give calculation of welds for K, X, Y, N joints.

So, can I take smaller value of ''a'' when designing RHS welded joints (like a=0.7t)? Can somebody give me guide, worked example of weld design for K, X, Y, N joints?

Share this post


Link to post
Share on other sites

Typical HSS grade in these environs is 350 MPa... likely similar to S355... not familiar with that designation. I don't know if you can get a copy of the text by Jeff Packer on HSS design... I'm not sure of the title... my copy's at the office... one of the best books in this area... Jeff has done a lot of work for CIDECT... and you may be able to source a copy through them.

Dik

Share this post


Link to post
Share on other sites

Is there any Eurocode equivalent to American AWS codes, which require that your leg size for a fillet weld on the end of material 1/4" thick and thicker be 1/16" less than the material thickness?

I've read about a nominal 2mm rule....but can't find the actual reference for it.

post-3046-14570089659102_thumb.jpg

Share this post


Link to post
Share on other sites

Thank you.

I had seen that thread, but it does not talk about what--if any--similar requirements exist in the Eurocodes, to match that '1/16" less than the thickness of plates 1/4" or thicker' approach called for in the American AWS codes.

I understand the reasoning behind the restriction, but--again--codes may set raw logic and pure physical laws as their touchstone* but the actual codes and their implementation....not so pure an embodiment of those worthy virtues.

* Yes, I'm being facetious

Share this post


Link to post
Share on other sites

From that "other" thread:

I think that there's a confusion between physical and effective throat sizes.

If I am trying to lap two 8mm thick plates, I could--theoretically, at least--place an 8mm fillet weld at the edge of plate"A" tying it to the side of plate "B", and vice versa from "B" to "A."

That 8mm leg gives me an 8mm*0.7071 = 5.7mm effective throat, PER WELD, or an 11.3mm combined effective throat (if I were, let's say, applying tension from the free end of "A" across the free end of "B").

HOWEVER, if I am making those welds to AWS D1.1, I would only be allowed to make (or consider) a weld with a 6mm leg (2mm less than the 8mm thickness of the plate, for plates over 6mm in thickness).

In ENGLISH: if I am making those welds to AWS D1.1, I would only be allowed to make (or consider) a weld with a .25" leg (1/16" less than the .315" thickness of the plate, for plates over .25" in thickness).

Share this post


Link to post
Share on other sites
Thank you.

I had seen that thread, but it does not talk about what--if any--similar requirements exist in the Eurocodes, to match that '1/16" less than the thickness of plates 1/4" or thicker' approach called for in the American AWS codes.

I understand the reasoning behind the restriction, but--again--codes may set raw logic and pure physical laws as their touchstone* but the actual codes and their implementation....not so pure an embodiment of those worthy virtues.

* Yes, I'm being facetious

Oddly enough the Eurocodes don't mention any kind of throat size restriction. If the American Codes provide such limits I think they should be physically valid (I mean if they work in America I'm sure they work in Europe too).

Actually, I think this problem is governed more by empirical data and in loco observation than by pure math...

Share this post


Link to post
Share on other sites

Agreed; I typically see the reasoning behind the US restriction as an issue of how to inspect, rather than anything else.

Share this post


Link to post
Share on other sites

Ok guys... lets make this more awkward :)

One of the PhD's at my university, who is involved with making of national annexes here in Latvia also thought of good practice and wanted to write a line in the national annex that says: "effective throat thickness should never be greater than 0,7*t(min)"... until we came upon this, and understood that EC3 basicaly says that every throat (in welds of hollow sections) should be thicker than the weaker material connected [EN 1993-1-8 7.3.1(4)]:

The design resistance of the weld, per unit length of perimeter of a brace member, should not normally

be less than the design resistance of the cross-section of that member per unit length of perimeter.

Maybe there is perfect description of what is brace member, but in most cases that is every hollow section member...

What do you guys think?

Share this post


Link to post
Share on other sites
Ok guys... lets make this more awkward :)

One of the PhD's at my university, who is involved with making of national annexes here in Latvia also thought of good practice and wanted to write a line in the national annex that says: "effective throat thickness should never be greater than 0,7*t(min)"... until we came upon this, and understood that EC3 basicaly says that every throat (in welds of hollow sections) should be thicker than the weaker material connected [EN 1993-1-8 7.3.1(4)]:

Maybe there is perfect description of what is brace member, but in most cases that is every hollow section member...

What do you guys think?

Hi! Why does this mean that the material should be thicker?

Share this post


Link to post
Share on other sites

Hallo!

It means that the throat should be thicker than the thickness of the weaker member. Not the members material thicker than the throat of the weld (a>t not a(max)<0,7*t or something like that).

Considering axial force (in this example tension), and lets say SHS with t = 4mm. The member is S355, so the resistance for unit length of perimeter is 355MPa*4mm=1420 N/mm. Now consider resistance of, lets say, fillet weld. We calcualte fillet weld as beiing the same material as the weakest element of connection (lets assume that our SHS member is the weakest and its S355). As this fillet weld is in a 90 degree fillet, the axial force on the member is transverse force on the weld, where the throat plane is on a 45 degree angle to the aplied force, so K = 1,225 [sCI Blue Book]. fu=490MPa for strongest S355 class, so fvw,d=490/(30.5*1,25*0,9)=251,5MPa. To get the resistence per unit length to transverse force we multiply it by 1,225, which then is = 308,09MPa - which means that the allowable stress from transverse force is less that that for the member of 355MPa. So, to get at least the same resistance to this axial force we need a throat of a=1420N/mm / 308,09MPa=4,61mm -> 5mm. Which as we see, is way more thatn 0,7*t of the thinest material.

Hope I'm not wrong in any way ;) That how we see it. So, basicaly, everytime you design hollow section brace member (again, not sure what is ment by brace member), you can not use the concept of weld throat being not more that 0,7*t of the thinest material.

Edited by ReinisGailitis

Share this post


Link to post
Share on other sites

"3.9 Design of welds

Except for certain K and N joints with partially overlapped brace members (as noted below), a

welded connection should be established around the entire perimeter of a brace member by means

of a butt weld, a fillet weld, or a combination of the two. Fillet welds which are automatically

prequalified for any brace member loads should be designed to give a resistance that is not less

than the brace member capacity. According to Eurocode 3 (CEN, 2005b), this results in the

following minimum throat thickness „‟a‟‟ for fillet welds around brace members, assuming matched

electrodes and ISO steel grades (IIW, 2008):

a >= 0.92 t, for S235 (fyi = 235 N/mm^2)

a >= 0.96 t, for S275 (fyi = 235 N/mm^2)

a >= 1.10 t, for S355 (fyi = 355 N/mm^2)

a >= 1.42 t, for S420 (fyi = 420 N/mm^2)

a >= 1.48 t, for S460 (fyi = 460 N/mm^2)"

That is a quote from "DESIGN GUIDE FOR CIRCULAR HOLLOW SECTION (CHS) JOINTS UNDER PREDOMINANTLY STATIC LOADING" from J. Wardenier ... , p.24

I think it is only for hollow sections with one sided fillet weld, and the rest for let say plates and Hot Rolled section is 0.7*t for one sided and 0.5*t for two sided fillet welds.

Share this post


Link to post
Share on other sites

Good post! Also the provided values seem to be reasonable, as my example above suggests a=1,15*t for S355 (for that example).

I also think that 0,7*t could be used for everything but hollow sections.

Hope these few last posts will shed some light on the subject for everyone using eurocodes :)

Share this post


Link to post
Share on other sites

Nice posts! I'm glad we got some solid conclusions.

Share this post


Link to post
Share on other sites

Hello, I have one doubt.....

Lets say I want to make T joint both side fillet welded from 8 and 16 mm thick plates ;

now I can make 2 arrangement:

1) 16 mm plate horizontal and 8 mm plate vertical (both side weld to 8 mm plate)

2) 8 mm plate horizontal and 16 mm plate vertical (both side weld to 16 mm plate)

so for both cases (both side) fillet leg size will be 6 mm (0.5*8/0.7)

or it will be 6 mm(0.5*8/0.7) for case 1 and 8 mm (0.5*16/0.7=11.43 mm but smaller plate being 8 mm thick) for case 2 ???

I am a bit confused? 

 

 

Share this post


Link to post
Share on other sites

http://fusionpoint.be/temp/equivalente keelhoogte scan.pdf

even though it's in dutch, should be useful.

 

This shows where the .92 ; .96; ... comes from.

Before one applies these formules, make sure you thoroughly understand the preconditions and limitations that comes with each situation.

If I ever get the time to translate this into English, I'll repost it again.

1 person likes this

Share this post


Link to post
Share on other sites

Dutch

Any way thank you 

I will try it with google Dutch to Eng translator.

Share this post


Link to post
Share on other sites

Did the above document help you in any way?

I always appreciate any feedback: If I could improve or clarify something in that doc, please let me know!

Share this post


Link to post
Share on other sites

Thanks for the document.

When we calculate same for S275 with "4.5.3.3 Simplified method for design resistance of fillet weld".

The k factor comes out to be 0.58.

What is your opinion.

Also as per EN 1993-1-8 4.1.(1)

"The provisions in this section apply to weldable structural steels conforming to EN 1993-1-1 and to material thicknesses of 4 mm and over. "

"For welds in thinner material reference should be made to EN 1993 part 1.3."

So I think table 4.1 of your doc should revised for the same.

Regards 

Gurudev

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now

  • Recently Browsing   0 members

    No registered users viewing this page.

  • Similar Content

    • By steven706
      Hi All,
       
      Do you have the design example for the Chevron Bracing according to EC3? or any book reference?
       
       
    • By Ee Hoe
      I am currently doing a beam strengthening by adding a steel plate to the bottom flange of my existing beam. But coming to the reduction of the fy, I realised that I am not very clear about the nominal thickness of the element.
      Is it the thickness of the flange or the web? Or is it the larger between 2?
       
      Hope to hear from you guys. Thanks alot!

    • By Hang
      Hello Engineers,  I am a Bachelor degree student in civil engineering. Do you guys have any workout pdf examples of steel beam design according to EC3. I am not familiar with EC3 so it could be nice to see some detail workout example. Thanks
    • By Nick Dunas
      We have finally launched!
       

       
      After two years of intensive work, we are more than happy to announce the initial release of the beta version of Steel suite.
       
      Steel suite is a cloud-based platform that offers high-end web applications for the design of steel connections according to the Eurocode. The initial version of the platform offers three applications for the design of fin plate simple joints. Additionally, an interactive steel sections library is also available.
       

       
      The main advantage of Steel suite is that the calculations are implemented on our servers. This feature assures that the engineer can access and use the applications from any device (PCs, laptops, tablets and mobiles) with an internet connection, regardless its computational power. The applications are developed using state of the art technologies, offering interactive 3D and 2D graphics.
       

       
      Please feel free to create a personal account and test the platform by yourself here:
       
      steel.ecwix.com/register
       
      You can navigate to our multilingual commercial web page here:
       
      www.ecwix.com/en
       
      If you find our project interesting keep in touch with us by following our page on LinkedIn:
       
      Ecwix Engineering @ LinkedIn
    • By Stelios89
      Hello everyone,
       
      I am not familiar with the use of forums so I am not sure how we can help each other but for starters I'd like some assistance with the design of a steel tee beam.
       
      The section is just shy of 0.5m, and on one side it is bearing onto masonry (100mm blockwork) and the other is bolted onto the web of a steel I-beam. It is also loaded on both its sides by approximately 5.5kN/m (total). 
       
      Please, if possible, provide some guidance as to how this section is classified and which parts of the eurocode would allow me to design it. I've already done some parts of it, where I basically assume that the section is in tension but I am not sure if I am missing anything.