Why scarf joints are no stronger than square joints

แชร์
ฝัง
  • เผยแพร่เมื่อ 29 พ.ย. 2024

ความคิดเห็น • 37

  • @LonersGuide
    @LonersGuide 10 หลายเดือนก่อน +5

    Generally, weldments are less ductile than the base material. Also, welding introduces stresses as the weld bead cools and shrinks. What some of us are attempting to avoid is a failure of the weld or its adjacent material due to these factors. Imagine flexing and twisting a tube with stiff spot that depending upon the material, may have extra ductile areas adjacent to it (HAZ). Sooner or later, either the weld will crack or the material immediately adjacent to it will. Scarfing a joint is an attempt to spread this difference in flexibility laterally in order to mitigate it. I have an old commercial welding textbook somewhere which indicates that if it is necessary to repair a ( medium or heavy commercial) truck frame, that a rectangular reinforcing plate the full height of the frame rail and something like twice as long should be placed over the repair, and welded only along the top and bottom (laterally), and never on the ends, as the perpendicular welds will eventually cause the frame to crack and fail. Many of us in the off-road fabrication scene will make perpendicular (square) joints in the frame tubing, but reinforce them with "fish" plates, the simplest form of which cover the joint like a diamond shape. Besides strengthening, fish plates reduce stress concentrations.

  • @a760541
    @a760541 2 ปีที่แล้ว +4

    The only guy i know who repairs his LR with a slide-rule!!!

  • @erictyrrell3560
    @erictyrrell3560 2 ปีที่แล้ว +2

    As the wrong type of engineer I have little understanding of your analysis but can appreciate the expertise presented. It would be interesting to demonstrate your position by welding scrap box tube together with several joining strategies and comparing their failure points.

    • @defendermodsandtravels
      @defendermodsandtravels  2 ปีที่แล้ว +3

      Those types of tests (static loading of conventional welded joints) were performed back in the forties and fifties and it'd be hard to dig out the reports now. Attention moved on to different things - the buckling of box girders, fatigue of all types of materials and structures, fracture toughness, buckling of columns, then tubular joints for offshore structures. The list goes on. I used to work in a structures lab in the early days of N. Sea oil (tubular joints and grouted connections).
      The early research work was incorporated in design codes and analysis methods which have stood the test of time so no one even questions them now. Modern design is mainly performed by software and many younger engineers probably only have a vague idea about the underlying equations.

  • @charlienzcharlie
    @charlienzcharlie 2 ปีที่แล้ว +1

    Hello, structural fabricator/welder here from New Zealand. We always use the square butt method, joint weld preparation is critical, these days we insert a backing strip to ensure ease of a sound full penetration weld. Ultrasound tested if the design engineer insists.

    • @defendermodsandtravels
      @defendermodsandtravels  2 ปีที่แล้ว +1

      You are doing it the correct way and you speak language I understand. Keep up the good work.

    • @defendermodsandtravels
      @defendermodsandtravels  ปีที่แล้ว

      If you speak the same language you must be a fellow engineer. Welcome friend.

    • @omribornstein8329
      @omribornstein8329 ปีที่แล้ว

      ...although....when adding any material for reinforment, you change the profile's moment of enertia.
      And you do not want to do that abruptly, thus sending you back to the elongated diagonal shape.
      Which was probably the main reason for all this confusion to start with....😂

  • @husq2100
    @husq2100 2 ปีที่แล้ว +1

    Thanks for your time , knowledge and effort in this series. The maths is over my head but it has given me a better understanding.

    • @defendermodsandtravels
      @defendermodsandtravels  2 ปีที่แล้ว +2

      The mathematical details are really there for other engineers (which is why this video is in the Technical Series playlist). If you know that the theoretical back up is there that's all that matters.

    • @husq2100
      @husq2100 2 ปีที่แล้ว

      @@defendermodsandtravels I walked through Mazda this morning, they had a new BT50 cab chassis in the yard. There were vertical welds (overlapping) of the main chassis rails. I took some photos for you if there is somewhere I can send them.

  • @omribornstein8329
    @omribornstein8329 ปีที่แล้ว +1

    I am not pretending to fully understand the math and physics behind your explanations, as I wasn't pay the needed attention when this stuff was taught...however I do want to suggest an optional root reason for this debate or misunderstanding, as you show it:
    A. The syress liads inflicted on a vehicle chassis (especially an offroad one) are not limited to bending and stretching stresses.
    The torsional stress loads may change your diagrams a bit.
    Another optional root cause for the "misunderstanding " may be the fact that whenever anyone elongates a longitudinal load bearing chassis member they always (should) back it up with an additional reinforment plate (at least according to most vehicle manufacturers) which will alter, locally, your chassis moment of enertia.
    Whenever yiu change your moment of enertia, for a given profile, you always wish to do it graduate manner.
    This prevent stress concentration and mainly is a good practice to prevent "hot" spots and reduce the chances for fatigue related, premature failure.
    I may be wrong about this but will still continue to do it this way.
    Very good video and clear explanations.
    Will continue watching the channel
    Thabk you

    • @defendermodsandtravels
      @defendermodsandtravels  ปีที่แล้ว +2

      Thanks for your contribution.
      There will indeed be torsional stresses present but these will be of lower magnitude than the bending and shear stresses shown here, particularly for closed members (e.g. box sections). However to calculate them I'd need to use a 3D model rather than the simple 2D analysis used here. This would inevitably involve a computerised analysis with a whole range of load cases. It'd be more visual but would be a whole lot more work for me and it'd be much harder to explain the results and conclusions in a concise manner, and they wouldn't be so very different. The simplified analysis I showed is what would be done at the conceptual stage before resorting to the computer and in days past would have been the only analysis performed.
      It isn't bad practice to use reinforcing plates particularly if they are well detailed and fabricated, however they should be applied where they are needed (i.e. in the flanges and not the web). In the structural fabrication industry they are used sparingly because of cost; The preferred approach is to make a straight cut and make a competent butt weld, job done.

    • @omribornstein8329
      @omribornstein8329 ปีที่แล้ว +1

      Thank you for the detailed reply, the info makes a lot of sense. Only reason I am raising it is that the vehicle manufacturers not allowing to do anything to the upper and lower sections and only have instructions of how to add reinforment while connecting to the web, even if the plates are bent and reinforce at least one of the flanges...
      So, still intrigued but understand that you are not my private teacher.😢
      Will be very interested to see an episode where you inspect a major vehicles manufacturer 's unfitter instructions and point out their misleading or misunderstood instructions.
      Keep up the good work and educational contribution
      Thank you

    • @defendermodsandtravels
      @defendermodsandtravels  ปีที่แล้ว +2

      @@omribornstein8329 I am unfamiliar with instructions issued by vehicle manufacturers however I am a Chartered Structural Engineer (equivalent to PE in US) and what I say on this subject is sound. For legal reasons I must make it clear I am not issuing advice to anyone.

  • @NathanNostaw
    @NathanNostaw 2 ปีที่แล้ว +3

    I really like these myth busting breakdowns. Thanks for another good video.
    The theory of this one went so far over my head I think the maths papers were taped to a satellite. I think I understood the final result though.
    Maybe the scarf joint is better when one does back yard chicken shit looking welds, that way the 10% of actual weld connection is more likely. :)

    • @defendermodsandtravels
      @defendermodsandtravels  2 ปีที่แล้ว +1

      So long as you got the gist of the video that's fine. The next few will be less technical I promise.

  • @rockcrawler31
    @rockcrawler31 10 หลายเดือนก่อน

    It would appear to me that perhaps a mistake that he's made, or assumes is that any chassis extension would be done immediately adjacent to a spring hanger where bending and rotational forces around the centroid are at their minimum in the overall view of a chassis' stress profile, and are shared somewhat equally with shear. Since there's no perfect world sometimes a chassis would dictate a join right in the middle of the frame at the peak of the tensile/compression forces below and above the centroid. If we assume the centroid ends up being a rotation pivot for these forces then perhaps a slanted scarf joint would place the weld more perpendicular to the actual tensile forces trying to seperate the joint.

    • @joeg7755
      @joeg7755 8 หลายเดือนก่อน +1

      FYI, and that is why all chassis OEM's recommend butt joints and to be completed around 6" before spring hanger. That will also allow enough backing plate on either side of weld.

  • @johnallen3555
    @johnallen3555 ปีที่แล้ว

    NNNOOO, please not your last word on the subject ! :) (it helps keep us all sharp)
    This is a really interesting and complex area to talk about. Your math is correct to the limits of the theory. In all engineering there is more at play than the bare numbers.
    ((really sorrry for the length of my reply, it is hard to put into words))
    My thoughts. (having been involved with the design and repair of many structural member failure repairs )
    - Virtually all welds put in some sort of stress concentration that would likely not be present in a single piece parent memember
    - It is undesirable to have said stress concentrations running at right angles to any tensile stress
    - in most situations of field repair, full and complete engineering analysis and calculations are not done. More often some astute estimates are made based on material knowlege, the life to failure ( number of cycles sort of thing) and the fundamental geometry of the item to be repaired and some basic calculations or sizing proportions.
    - Structural welds done in te field ( even by expert welders) should be given the very best of opportunities to be as insensitive to weld defects, material differences and stress concentrations as possible
    If the failure is life / fatigue related (as opposed to impact damage ),
    - at best an unreinforced repair will achieve similar, but somewhat lesser life that the original failed structural member
    - We usually aspire to do better than this replication of inadequacy (but some times emergency repairs, time constraints or a non stratigic panicked client dictates otherwise)
    - for field repairs a weld ( or any non reinforced repair) must be considered a weak point
    Now as things fail, the discussion now moves to how they fail (lets keep with the fatigue / over stress theme for discussion), the issue becomes one of the nature and speed of the failure.
    - Stress concentrations lined up like a zipper at right angles to the principle stress are like the peforations in toilet paper (yes they put the stress raiser (peforations) across deliberatly because the paper will most likely tear on the line)
    We don't want to do this toilet paper thing with a weld repair so, what are the options
    - more often than not, the paper weld repair is done and the joint is plated over with a hopefully designed plate that places its welds and stress concentrations in a manner so as to improve to overall member and new joint service life without adding further problems (yup this is easily done if not carefull, too much metal, changed stiffness in a dynamic flexable structure can cause problems)
    Now I do recognise that qualiyt controlled factory weld process can deliver transverse butt welds of the full strengh (done all the time, and sucessfully too by major manufacturers such as Catipillar).
    - for less controlled welding situations where an item has demonstrated its structural limitations / inadequacy over time, the metal aint what it used to be when it left the factory
    - any plating over the repair must be shaped to connect in a mannerthat does not put welds, their defects roughness and the geometric stress raisers where a crack would naturally want to run
    As a side note to structural repairs on dynamic structures, surface finish and geometry of plating can be really important
    - tapering down of plating to reduce stress raiser makes a big difference when your back is to the wall
    - one can not effectivly taper a square doubler plate with a transverse weld to negate geometry stress raisers as the transverse fillet weld is ground too. So, what to do
    - What to do is to put the weld at an angle (yup shape the plate, weld past the plate ends and taper weld ends and plate off with post fabrication grinding, being carefull to rund the grind marks in line with the principle stresses

    • @defendermodsandtravels
      @defendermodsandtravels  ปีที่แล้ว +1

      I will give your reply proper consideration since it obviously deserves that. I will post a reply in due course when I have read the post properly.
      My background? PhD in Fracture Mechanics. My supervisor was Paul Paris (of Paris' law for FCG). 40 years in the offshore industry where fatigue and fracture are primary design considerations. Sometime expert witness and part time lecturer in FM at IC. If I find we are greatly at odds it'll be interesting but I am always prepared to be proved wrong :-) I am not actually arrogant although well qualified in the subject matter.
      I send this from a bar in Borneo overlooking the South China sea. It's tough here :-)

    • @defendermodsandtravels
      @defendermodsandtravels  ปีที่แล้ว +1

      Having quickly scanned your reply..
      The highest stresses in beams are in the top / bottom flanges and that's where the failures arise. Scarf joints in the webs are irrelevant but they keep the YT dudes happy.
      I have never seen angled welds in flanges - this would give mixed Mode I / II cracking which is poorly understood.
      You can get good fatigue performance from butt welds under the right conditions. If you have access to both sides then no problems. If single sided (as is common with tubular structures, my field) you must have close control over the root pass (usually TIG) and the fill runs are less critical.
      If you can't do a high quality butt weld then doubler plates are a good alternative, profiled to avoid sharp transitions as you say. Good weld profiles are necessary in fatigue critical situations, ideally with profile grinding and shot peening. All of this done in industry however the average YT dude wouldn't bother.
      To put it in perspective, my hand calcs on the Discovery 1 chassis rails showed that they aren't fatigue critical. However on my recent trip to Namibia I met a couple who managed to break one of of chassis rails on their Hilux. This was only 2/3 the depth of the LR equivalents and I don't need to explain the significance of this.
      I hope we are on the same wavelength.

    • @johnallen3555
      @johnallen3555 ปีที่แล้ว

      @@defendermodsandtravels
      yes we are on the same wave length
      - I really appriciate your videos
      - as a young mech engr in the 80s I was exposed to an awefull lot of cracked frames and structures in the rail industry (rail maintenance machinery was a particularly crackfull learning enviornment, some of the machines in the field were cracking daily / weekly, yes major frame cracks had got out of hand).. Attention to detail, and designed repairs (lot of learning as to what worked and what did not!)
      - good welding, removal of the overstress aged materials, plating designed to minimise stress concentration from the plate connection geometry and avioding the toilet paper thing in the geometry
      - these days I use Strand 7 FEA, such programs are very good at modeling the impact of geometry , welds etc if one has the time or inclination to do so ( I am semi retired and do have the time for such things!)
      any way thanks for your videos,
      they are really good

    • @defendermodsandtravels
      @defendermodsandtravels  ปีที่แล้ว

      @@johnallen3555 Pleased we are on the same wavelength and not at cross purposes.
      We probably have broadly similar experiences albeit in different industries.
      Something clearly went wrong with UK rail in the early 2000s although it seems to gave calmed down now

  • @gkizzle1265
    @gkizzle1265 10 หลายเดือนก่อน

    this man is truly a brain, genius one might say

  • @gavinralph2910
    @gavinralph2910 2 ปีที่แล้ว +2

    scarf joints are best reserved for wood😄, when you want the bonded surface areas as large as possible to ensure a uniformed bond strength,
    Works well on boats...not so great on Land Rover chassis🤣

    • @defendermodsandtravels
      @defendermodsandtravels  2 ปีที่แล้ว +1

      Agreed. You certainly seem to have picked up the message of the video. With glued timber the joint is a plane of weakness but not so with butt welded steel.

  • @mk109siii9
    @mk109siii9 2 ปีที่แล้ว

    Hi Bill. I want to ask you a Land Rover related engineering question, I can't find any contact info, so please excuse me for posting my question in the comments section.
    It relates to Oil Pressure drop in a 300tdi. When the engine gets warm the oil pressure drops from 45psi to 15psi, if I coast downhill and drop the engine temp below 74.C the pressure rises again to about 25psi. I doubt that a temp variance of

    • @defendermodsandtravels
      @defendermodsandtravels  2 ปีที่แล้ว

      Michael, whilst I'd like to help I think it'd be unwise for me to get involved in mods to your vehicle's lubrication system so I will just give general answers. In response to your question it is possible to calculate the oil flow / pressure drop but you really need to model the whole system and that's quite a bit of work and requires good input data and a lot of experience. It's what Piping Engineers do for a living. If you want to calculate the flow through a washer in isolation then do a search on "orifice plates" which is what they are called in the piping world.
      Provided your oil pump is delivering as designed, the clearances are as per spec and you are using the corrrect grade of oil you should be getting full pressure - that's what the physics says. Most people don't have an oil pressure gauge so don't know how low it gets when hot. The Daihatsu engine in my Defender starts at nearly 6 bar when cold and can reduce to 1.5 bar when really hot which is a bit concerning but it seems to be all right with that. However 1 bar would worry me (the low pressure light must be close to coming on) and I agree you need to look into it.

    • @mk109siii9
      @mk109siii9 2 ปีที่แล้ว

      @@defendermodsandtravels Hi Bill. Thank you for your response and guidance to a formula. Having searched the net, the issue seems to be widespread among 300Tdis. Prior to the rebuild I had around 10psi more pressure at start up and at hot running. That was with a later Brazilian made MWM block, which may offer more resistance in the oil galleries (I don't know) and with the gauge sender unit bolted to the end of the T-piece adaptor, now it's at 90. angle. I don't think this should make a difference.
      I do have a custom radiator (triple flow with 30% increased capacity) and I did notice a mild drop in oil pressure when I first fitted that (difference in resistance between that an the old radiator). Looking at Mike's videos on Britannica Restorations (TH-cam) the oil cooler circuit never gets 100% of the oil flow, even when fully open, so that 'hot' and 'cold' oil mix at the filter. Given that with a "tight restrictor" ie: thermostat closed, everything runs fine, and that with "no restrictor" (thermostat open), the oil still bifurcates at the filter, then I think creating a 'partial restrictor' should not affect oil flow though the filter and to the engine. The restriction I'm contemplating is in a one-way circuit (filter/thermostat housing - cooler - filter/thermostat housing) so if I offer resistance inside that circuit, am I achieving anything?
      Given my ignorance, I'm happy to step back from this contemplation if you can kindly show me any risks I haven't thought of.
      Thank You, Michael

    • @defendermodsandtravels
      @defendermodsandtravels  2 ปีที่แล้ว

      @@mk109siii9 Michael I can't really offer much guidance because I don't know the 300Tdi lubrication system in detail, and I don't want to get involved in something I don't know about. I suggest you find other owners who have had the same problems and share your experiences with them. Good luck.

  • @stephanc7192
    @stephanc7192 6 หลายเดือนก่อน

    Great video

  • @adelarsen9776
    @adelarsen9776 2 ปีที่แล้ว +1

    That'll do me :-)
    Really the situation is education. Not everyone had the benefit of a science Grammar school education where most of what you said was taught and tested in 2nd and 3rd Form physics.
    What we are seeing in real time is "Barn Door" engineering. Near enough is good enough.
    Well it's not good enough if you're on Piper Alpha or Alexander Kjelland.

    • @defendermodsandtravels
      @defendermodsandtravels  2 ปีที่แล้ว +3

      I was personally involved in the aftermath of Piper and used to see the upturned hull of the Kielland in Gandsfjord on frequent visits to Stavanger in the 1980s. My biggest lesson from that, as an engineer, was to have a constantly enquiring mind and to challenge all assumptions. Barn Door engineering is just the opposite - you just know you are right and have contempt for anyone who questions you.

    • @adelarsen9776
      @adelarsen9776 2 ปีที่แล้ว

      @@defendermodsandtravels Indeed.
      Happy Days :-)

  • @garypoulton7311
    @garypoulton7311 2 ปีที่แล้ว

    Surely the loading is vertikal, on a chassis connection, rather than horizontal? Not saying you are, I really dont know. Iimagine the loading on a chassis is vertical, and twisting motion.

    • @defendermodsandtravels
      @defendermodsandtravels  2 ปีที่แล้ว +1

      I have no problems with anyone questioning my videos since I could always make a mistake.
      You are correct that the loading on the top of a chassis member is lateral however the local bearing stresses are at least an order of magnitude lower than the bending stresses. It may not seem logical but that's the way it is.