The Yard Guys

[jimbunyard]

As long as the engine year and manufacturer were the same or close I would agree. The weight difference of modern engines at any rate would be negligible, as would the 25 hp increase. The owners mistake was, in effect, adding 6 inches to an already inadequate transom by using a jack plate, thereby increasing leverage exactly where it could do the most harm.

[neilpride]

Originally Posted by jimbunyard

Man, that ITW video was disheartening. I guess those glue dispensers feel they're lucky just to have a job (or maybe they're Deltas and Epsilons). I imagine their days are numbered though, the corprocats will soon give those jobs to robots...

The most interesting thing about this picture;



in comparison with this picture;


is the difference in the size of the radii in the fillets. The fillet in the Plexus joint is easily twice that of the fillet in the tabbed one. Does this indicate a testers' bias, or inconsistent experimental controls, or designed joint criteria? It makes a huge difference.

Additionally, from what I know of Plexus's almost crystalline hardness (I'm aware that ITW makes at least dozens of products and that some of them should include flexibility), the use of a deformable core panel in the test leaves me with more questions than answers, as regards this particular application.

I prefer bonding, filleting and tabbing. Hope I'll be forgiven, the pictures below are from a 21 ft., 85 mile an hour bass boat, but the idea is the same, When does use become abuse?, and seems in line with the general question about what makes a boat seaworthy (dare I say bluewatery?)

The boat was rated for 200 hp, the owner put a 225 on it, which is not uncommon. He also added a 6 inch jack plate, which resulted in what you see in the pictures. The buttresses (for lack of a better word, knees maybe) were tabbed but not bonded and were still securely attached at the floor. The boat itself when getting on plane pulled severely to port and once on plane listed about four inches down to port while running, indicating an induced hull twist under power.

The first picture is just an overview showing what is to me a basic design flaw, an unsupported transom.

2 and 3 are closeups of delaminated tabbing.

Picture 4 shows what is to me the correct way of installing bulkheads (although in this case they're actually stringers), that is; bonding the panel to the hull, filleting as generously as possible and practical, and tabbing, in this case with a layer of 18 oz roving 5" wide, followed by a layer 8" wide followed by a layer of .5 oz mat 10" wide, all laid wet one after the other.

Picture 5 shows the overall substructure, the different fillet radii where the stringer meets the transom on the outboard versus the inboard is the key point.

Picture 6 is just a vanity picture after gelcoating.

pictures 7 and 8 show the reason for the different radii of the fillets where the stringer meets the bulkhead. While I felt pretty sure the new reinforcements would be more than adequate, I also knew why the owner brought the boat in and therefore recommended the addition of the aluminum plate, just to CMA.

Pictures 9 and 10 are also vanity pictures, but I always like to see the end result. It's interesting though, in picture 10, the "hand laid hull" logo doesn't say anything about how strong the boat was, or wasn't. Often times the owner, or the conditions, have more to say about that. (The manufacturer is no longer in business)

And by the way, the white putty shown in the fillets is just glass beads mixed a little wet, with a little silica added to control runny-ness. My seat of the pants tests show that this putty is stronger than PVC foam or wood and is much easier to sand than milled fibers or anything else. This yields the smoothest fillet, which allows the best tab layup, which yields the most strength. It's the system, not the individual components, that give strength.

Which of course leads back to '(mass) production' boats. Many production boats cannot be assembled in any other way than using the glue in liner or modular approach, indeed, it appears that many boats are designed around the solution (a glue gun) rather than the other way around.

For a status symbol, no problem. To trust these boats in the Northern North Atlantic or South Polar Seas, I'd try to make damn sure the glue spreader was fat and happy....

LoL!!! Good one!! i miss it, the plexus fillet is thicker....

[Terra Nova]

Originally Posted by jimbunyard

...the white putty shown in the fillets is just glass beads mixed a little wet, with a little silica added to control runny-ness. My seat of the pants tests show that this putty is stronger than PVC foam or wood and is much easier to sand than milled fibers or anything else. This yields the smoothest fillet, which allows the best tab layup, which yields the most strength...

Jim--good eye and good post.

But glass beads make the weakest putty.

[smackdaddy]

Originally Posted by jimbunyard

Which of course leads back to '(mass) production' boats. Many production boats cannot be assembled in any other way than using the glue in liner or modular approach, indeed, it appears that many boats are designed around the solution (a glue gun) rather than the other way around.

For a status symbol, no problem. To trust these boats in the Northern North Atlantic or South Polar Seas, I'd try to make damn sure the glue spreader was fat and happy....

I'm not sure who's recommending mass production boats for voyaging in the Northern North Atlantic or South Polar Seas - but I sure wouldn't...even though one or two have done it.

[smackdaddy]

Originally Posted by jimbunyard

The most interesting thing about this picture;



in comparison with this picture;


is the difference in the size of the radii in the fillets. The fillet in the Plexus joint is easily twice that of the fillet in the tabbed one. Does this indicate a testers' bias, or inconsistent experimental controls, or designed joint criteria? It makes a huge difference.

I don't see a bias in that thesis report. Starting on page 83 (and going for next 15 or so pages) you can see where they came up with the sizing for the tabbing fillet. It's from recommendations from various studies by the Gougeon Brothers Company, R.A. Shenoi and F.L.M. Violette, Burchardt, etc.:

According to Gougeon Brothers the main way of connecting an out-ofplane joint is by means of a fillet. The materials chosen were WEST SystemR epoxy-based resin and, according to the purpose, 406 Coloidal Silica or 409 Microsheres fillers. For structural joins 406 Coloidal Silica was recommended, resulting in a smaller fillet as is seen in Fig. 4.1.

A recommendation of the test to be performed, the materials and the radius of the fillet used are shown in Fig. 4.2. Here a force is applied until the destruction of the joint, at a distance equal with eight times the thickness of the material being joined. If the breaks occur in the piece being joined, it was assumed that a sufficient joint strength has been achieved. If the break occurs at the interface between the fillet and the part, or inside the fillet, it was advised to increase the radius of the fillet or to change the properties of
the filling material. No recommendations were made about the way the sample was to be supported.

Then the specifics of the actual test radii and why on page 92:

In the case of tabbed T-joint (traditional joint), a R12mm radiused fillet of Q-Cel/vinyl ester filler (same composition as described in section 2.13.4) has been added with the help of a radiused spatula. Then the first layer of tabbing was added and wetted throughout with resin with a brush. Then the second layer was laid down with the fibres at 90 ◦ with the first one, in this way the tabbing runs at +/−45 ◦ across the joint. The joint was left to cure over night before it was carefully cut with a skill saw. Initially, the panels were designed oversized (so the measurement for the three samples started from the middle of the panels) and the ends were discarded in order to eliminate any possible manufacturing imperfections at the end of the fillet.

In the vinyl ester/filler mix and Sikaflex 252 cases, a R40mm adhesive
radius was applied, and in Plexus MA550 case a R20mm adhesive radius was applied using a radiused spatula. In the first case, the radius was chosen similar to common practice [22] and [14] (which represents a recommended solution for high density fillets), while in the Sikaflex 252 case, the radius was chosen similar to common practice [14] (for low density fillets). Due to the high price of the Plexus product, and due to the high mechanical properties of the adhesive, a smaller radius was chosen. The cutting procedure was the same as in the case of tabbed joints.

Remember, this is a Master's Thesis from an Engineering Grad Student - he's not going to be winging it. And he's not going to be hawking one product over the other. This isn't a marketing piece.

So you guys really should read through it before jumping to conclusions.

[jimbunyard]

Doesn't really matter, the fact is that the strength of the joint is directly proportional to the radius of the fillet, whether it's made of milled fibers, fumed colloidal silica, phenolic beads, glass beads, or sawdust containing ground steel, wood, uhmw and aluminum from my table saw (which I've (obviously) tried and used successfully (only on my own projects though), and the quantity and quality of the laminate, if used.

As is stated in your own quote.



"Quote:
According to Gougeon Brothers the main way of connecting an out-ofplane joint is by means of a fillet. The materials chosen were WEST SystemR epoxy-based resin and, according to the purpose, 406 Coloidal Silica or 409 Microsheres fillers. For structural joins 406 Coloidal Silica was recommended, resulting in a smaller fillet as is seen in Fig. 4.1.

A recommendation of the test to be performed, the materials and the radius of the fillet used are shown in Fig. 4.2. Here a force is applied until the destruction of the joint, at a distance equal with eight times the thickness of the material being joined. If the breaks occur in the piece being joined, it was assumed that a sufficient joint strength has been achieved. If the break occurs at the interface between the fillet and the part, or inside the fillet, it was advised to increase the radius of the fillet or to change the properties of the filling material. No recommendations were made about the way the sample was to be supported."


Then we have this little jewel:


"In the case of tabbed T-joint (traditional joint), a R12mm radiused fillet of Q-Cel/vinyl ester filler (same composition as described in section 2.13.4) has been added with the help of a radiused spatula. Then the first layer of tabbing was added and wetted throughout with resin with a brush. Then the second layer was laid down with the fibres at 90 ◦ with the first one, in this way the tabbing runs at +/−45 ◦ across the joint. The joint was left to cure over night before it was carefully cut with a skill saw."


So the joint was left to cure overnight before cutting carefully with a Skil-Saw (must be accurate on these masters thesi). And then they tested it? or they let it cure fully? or they autoclaved it? or they put it in their car in the sun for three hours? and then tested it. And, to be fair, what about the Plexus? What's it's cure time to full strength, 2 hours, 2 days?

The fact that it's part of a Masters thesis lends (very marginal) credibility only if the applicant was successful in his application for a Masters Degree.

If the criteria is the strongest joint, then the material applications should be designed to yield the strongest joint, regardless of dimensions.

If the criteria is to find the strongest joint per unit size, then the experiment should compare joints with the same dimensions (as originally pointed out).

If the criteria is to test the strength of the joining material, then the experiment should be designed to test the strength of the material itself, not the material in combination with other variable components. (Also originally pointed out, although somewhat obliquely.)(Of course, that information could more easily be found by reading the spec sheet of the material in question.)

As is shown by Neil Pride's pictures of Bent Toes' application in the rudder thread (and probably others' pictures too, the thread's gotten too long to read through) the problem is not in the strength of plexus but in it's application (more properly, mis-application) and the over-reliance on it.

If you don't want people to jump to conclusions (and that is not an admission that I did, as evidenced by my original post), don't invite them to do so by posting silly pictures illustrating ill-defined assumptions like these:



"So, at the end of the day, Plexus was clearly the strongest method of adhesion - the the point that the flange materials (bulkhead and hull) break before the joint lets go.

Despite the delam shown in Neil's photo, I'm just not seeing much evidence out there that traditional tabbing is better overall than an adhesive fillet like Plexus."



So despite the real world evidence, you're not seeing much evidence?!?!

Kinda like me telling you your autopilot mount was plenty strong while you're showing me where the arm ripped off the base.

[smackdaddy]

Originally Posted by jimbunyard

Doesn't really matter, the fact is that the strength of the joint is directly proportional to the radius of the fillet, whether it's made of milled fibers, fumed colloidal silica, phenolic beads, glass beads, or sawdust containing ground steel, wood, uhmw and aluminum from my table saw (which I've (obviously) tried and used successfully (only on my own projects though), and the quantity and quality of the laminate, if used.

As is stated in your own quote.



"Quote:
According to Gougeon Brothers the main way of connecting an out-ofplane joint is by means of a fillet. The materials chosen were WEST SystemR epoxy-based resin and, according to the purpose, 406 Coloidal Silica or 409 Microsheres fillers. For structural joins 406 Coloidal Silica was recommended, resulting in a smaller fillet as is seen in Fig. 4.1.

A recommendation of the test to be performed, the materials and the radius of the fillet used are shown in Fig. 4.2. Here a force is applied until the destruction of the joint, at a distance equal with eight times the thickness of the material being joined. If the breaks occur in the piece being joined, it was assumed that a sufficient joint strength has been achieved. If the break occurs at the interface between the fillet and the part, or inside the fillet, it was advised to increase the radius of the fillet or to change the properties of the filling material. No recommendations were made about the way the sample was to be supported."


Then we have this little jewel:


"In the case of tabbed T-joint (traditional joint), a R12mm radiused fillet of Q-Cel/vinyl ester filler (same composition as described in section 2.13.4) has been added with the help of a radiused spatula. Then the first layer of tabbing was added and wetted throughout with resin with a brush. Then the second layer was laid down with the fibres at 90 ◦ with the first one, in this way the tabbing runs at +/−45 ◦ across the joint. The joint was left to cure over night before it was carefully cut with a skill saw."


So the joint was left to cure overnight before cutting carefully with a Skil-Saw (must be accurate on these masters thesi). And then they tested it? or they let it cure fully? or they autoclaved it? or they put it in their car in the sun for three hours? and then tested it. And, to be fair, what about the Plexus? What's it's cure time to full strength, 2 hours, 2 days?

The fact that it's part of a Masters thesis lends (very marginal) credibility only if the applicant was successful in his application for a Masters Degree.

If the criteria is the strongest joint, then the material applications should be designed to yield the strongest joint, regardless of dimensions.

If the criteria is to find the strongest joint per unit size, then the experiment should compare joints with the same dimensions (as originally pointed out).

If the criteria is to test the strength of the joining material, then the experiment should be designed to test the strength of the material itself, not the material in combination with other variable components. (Also originally pointed out, although somewhat obliquely.)(Of course, that information could more easily be found by reading the spec sheet of the material in question.)

As is shown by Neil Pride's pictures of Bent Toes' application in the rudder thread (and probably others' pictures too, the thread's gotten too long to read through) the problem is not in the strength of plexus but in it's application (more properly, mis-application) and the over-reliance on it.

If you don't want people to jump to conclusions (and that is not an admission that I did, as evidenced by my original post), don't invite them to do so by posting silly pictures illustrating ill-defined assumptions like these:



"So, at the end of the day, Plexus was clearly the strongest method of adhesion - the the point that the flange materials (bulkhead and hull) break before the joint lets go.

Despite the delam shown in Neil's photo, I'm just not seeing much evidence out there that traditional tabbing is better overall than an adhesive fillet like Plexus."



So despite the real world evidence, you're not seeing much evidence?!?!

Kinda like me telling you your autopilot mount was plenty strong while you're showing me where the arm ripped off the base.

Sooooo, did you actually read the thesis?

[jimbunyard]

Originally Posted by Terra Nova

Jim--good eye and good post.

But glass beads make the weakest putty.

You are correct, sir.

I should have included a more direct caveat than I did, stating that if one is making their own repairs they should stick to the recommendations of the manufacturers supplying the products being used, especially if they have no benchmark knowledge to go by.

I said that I used microballoons (glass beads) for two reasons; one as a kind of an offhand slap at smackdaddies dogmatism and two, somewhat unfairly, as a dig at amateurs' (me included) reliance on packaged instructions.

But there are two real world reasons I use beads instead of milled fibers, one supported by real world application, the other supported by real world inductive reasoning.

The beads smooth in a fillet more easily and uniformly and sand out more quickly, yielding a more uniform, smooth layup over them, with little or no entrapped air, which in turn yields a stronger joint.

Glass beads and resin, mixed a little wet to promote adhesion, in my experience are stronger than the materials (wood and pvc foam) that I usually use.
If I glue two pieces together and then pull them apart after curing, most of the delamination at the interface is matrix stuck to glue, but some (usually significantly less) of the glue is stuck to the matrix. Which leads me to believe that they are close in their mechanical properties (don't know the technical terms).

So my rationale is; if the mechanical properties in a mechanical system are closer, then the likelihood of inducing stress points is reduced, which in theory should lead to a more resilient rigid structure, compared to a mechanical system with unevenly distributed forces caused by using stress inducing components of varying mechanical properties.

Haven't had any failures that I know of yet (25 plus years) but have had several fin box blowouts on windsurfers using pure resin (both polyester and epoxy).

Seems kind of like being a good manager, I'd rather have 5 average people that can work together than 4 average people and one brilliant *******.

[Emmalina]

So how many stupid motor boaters put a 225 on a 200 rated boat. Wonder why they break :-/

[Emmalina]

Here's my dusty yard team... 220kg of epoxy 65kg and 2 days for 5 layers of woven on the hull !
Got them a few beers when we finished

[Terra Nova]

Jim--you must know what shortcuts you can get away with. For my own builds, I don't use microballoons except a small amount for final fairing, never under cloth, where the strongest bond will be produced with the glass directly against the substrate.

For typical tabbing, high density filler, or microfibers and colloidal silica, are what I use for filets which, when carefully applied and cleaned up, while wet, present no more difficulty to prepare for glassing, which can often then commence before the filets are completely cured. These filets add tremendous strength to the bonded joint, and can often stand alone. Otherwise the high strength putty provides a strong substrate to which subsequent fiberglass is bonded.

[Terra Nova]

xxxxxx

[jimbunyard]

Originally Posted by smackdaddy

Sooooo, did you actually read the thesis?

Why? Should I consider that a strawman question or an attempt to draw mw into a strawman argument?

I thought my opinion on the thesis was rather clear.

And by the way, I have to ask. Did you read the thesis? Or maybe I should ask did you understand what was written?


From the esteemed thesis:

My comments are in italics.

As shown, a purely theoretical estimate of joint strength is not possible or
acceptable as a basis for the design because of the uncertainty about imperfections, local stress concentrations and materials modes of failure. Without an expensive testing program, one can only approximate the stresses between the parts involved in the connection.

This is pretty vague. On one hand they seem to be saying one needs to use a real world estimate of joint strength in designing. On the other it seems they're saying that their own (and anyone else's) estimate of joint strength is not to be trusted because of lack of funds. And there are several other possible interpretations...


No attempts were made to calculate manufacturing prices, but it is expected
that the execution time will be shortened in the case of adhesives fillet. The
perspective of automatisation will make the joining process cheaper.

In this case 'cheaper' means weaker. See the following statements on strength.


No ageing effects or fatigue and creep effects have been taken into consideration during the testing program. For a full characterization of the adhesive adherend T-joint system these effects have to be studied.

So two of the main concerns we as consumers, and more specifically consumers who trust their lives (at least in part) to a product, are not even being considered.


In the case of a T-joint with tabbing and QCel microballoons fillet R12mm,
a graph was generated representing the load per unit width of joint (10mm)
(see Fig. 4.12). The average maximum load recorded was 44.8N/10mm, the
average deflection recorded at peak load was 25.6mm.

This tabbed joint required about 7N/10mm more force than the Plexus joint before what they considered failure.


In the case of a T-joint with vinyl ester/Epiglass HT120 fillet R40mm, the
graph representing the load per unit width of joint (10mm) is presented below
(see Fig. 4.13). The average maximum load recorded was 57.5N/10mm, the
average deflection recorded at peak load was 29.1mm.

This tabbed joint required about 20N/10mm more force than the Plexus joint before what they considered failure


In the case of a T-joint with Plexus MA550 fillet R20mm, the graph representing the load per unit width of joint (10mm) is presented in Fig. 4.15. The average maximum load recorded was 37.45N/10mm, the average deflection recorded at peak load was 32.47mm.

This Plexus joint apparently caused the failure of the panels by application of force either 7n/10mm or 20/N/mm earlier than the two tabbed joints. So the tabbed joints were roughly 25 and 75 percent stronger (That's what I get in my head, you figure it out) Not only that, the Plexus joint deflected further than either of the others, under substantially less load.




So to sum up, if I use the Plexus joint as described in the above thesis, what I get is a joint that deflects more, under less load, and is prone to inducing (potentially catastrophic) panel fracture, as compared to traditional fillet and tabbed joints.


But to answer your question, no I didn't read the entire thesis. The mathematics were way over my head (hence possibly redeeming I suppose), but most of the prose was pretty sophomoric, unclear and in-concise.

Kinda what I expected from the examples you posted.

But I did read the T-joint (and other sections) in depth.

So that leads us back to your assertions.

It could be construed that you have some hidden, unfathomable agenda.

It could be that you're a crypto-engineer, i.e. you have to believe the results to see them. (Or in this case misrepresent them)

Or you could (somewhat perversely) enjoy making other people use their brains.

Or any number of other postulates.

But in this, as shown by the thesis you cherry-picked, and the real world experience of many people (for me not against the Plexus but for the bonded, filleted and tabbed), the evidence is overwhelmingly against the idea that Plexus is better in these applications.

But hey, chose for yourself. That's part of the reason I entered this discussion with an example of failed tabbed joints.

[FamilyVan]

Yard guys must be more trustworthy some places than others.
I was having problems with my atomic 4. I wasn't sure what the problem was but I didn't think it was serious. I was working a lot and didn't have time to fiddle with it myself, so I had the marina look at it. It wasn't starting and I suspected the coil. The boat was up for sale, and I needed it running to sell it.
The yard guys looked at it that October and said they couldn't get it running it was a "serious" problem.
I paid for winter storage, spring came along and they couldn't get it running, it was a "serious" problem. Then summer came along, I still didn't have time to deal with the boat myself because I'm a sailor and when I work, I'm really not around. The boat was launched (the yard didn't have room to store it on land) and I was paying to keep the boat in the water. I was calling the marina twice a week and getting the same answers. They were working on it, it was a serious problem, they didn't have time for it but they'd look at it next week etc. I asked if I could bring in an outside mechanic- the answer of course was no insurance blah blah blah.
Then- my brother, who was working at the marina as a tender Captain over heard a conversation between a couple of the yard guys. They knew I was out of town, they knew it was a good boat in good condition, they knew I needed a new coil and plugs, they knew the boat was up for sale.
One of them (forgetting my brother was my brother) actually joked that he wasn't changing the coil and was playing me because he knew I wouldn't want to pay for another seasons storage and he was going to make a ridiculous low ball offer on the boat for himself!
My brother related this to me, I called in an outside mechanic (quietly) and he had the boat running in about 15 minutes with a new coil! I got the boat out of there to another marina ASAP and was able to sell the boat the following week for a fair price.
Sorry, I wouldn't trust the yard guys in an outhouse with a spoon.

[smackdaddy]

Originally Posted by jimbunyard

Why? Should I consider that a strawman question or an attempt to draw mw into a strawman argument?

I thought my opinion on the thesis was rather clear.

And by the way, I have to ask. Did you read the thesis? Or maybe I should ask did you understand what was written?


From the esteemed thesis:

My comments are in italics.

As shown, a purely theoretical estimate of joint strength is not possible or
acceptable as a basis for the design because of the uncertainty about imperfections, local stress concentrations and materials modes of failure. Without an expensive testing program, one can only approximate the stresses between the parts involved in the connection.

This is pretty vague. On one hand they seem to be saying one needs to use a real world estimate of joint strength in designing. On the other it seems they're saying that their own (and anyone else's) estimate of joint strength is not to be trusted because of lack of funds. And there are several other possible interpretations...


No attempts were made to calculate manufacturing prices, but it is expected
that the execution time will be shortened in the case of adhesives fillet. The
perspective of automatisation will make the joining process cheaper.

In this case 'cheaper' means weaker. See the following statements on strength.


No ageing effects or fatigue and creep effects have been taken into consideration during the testing program. For a full characterization of the adhesive adherend T-joint system these effects have to be studied.

So two of the main concerns we as consumers, and more specifically consumers who trust their lives (at least in part) to a product, are not even being considered.


In the case of a T-joint with tabbing and QCel microballoons fillet R12mm,
a graph was generated representing the load per unit width of joint (10mm)
(see Fig. 4.12). The average maximum load recorded was 44.8N/10mm, the
average deflection recorded at peak load was 25.6mm.

This tabbed joint required about 7N/10mm more force than the Plexus joint before what they considered failure.


In the case of a T-joint with vinyl ester/Epiglass HT120 fillet R40mm, the
graph representing the load per unit width of joint (10mm) is presented below
(see Fig. 4.13). The average maximum load recorded was 57.5N/10mm, the
average deflection recorded at peak load was 29.1mm.

This tabbed joint required about 20N/10mm more force than the Plexus joint before what they considered failure


In the case of a T-joint with Plexus MA550 fillet R20mm, the graph representing the load per unit width of joint (10mm) is presented in Fig. 4.15. The average maximum load recorded was 37.45N/10mm, the average deflection recorded at peak load was 32.47mm.

This Plexus joint apparently caused the failure of the panels by application of force either 7n/10mm or 20/N/mm earlier than the two tabbed joints. So the tabbed joints were roughly 25 and 75 percent stronger (That's what I get in my head, you figure it out) Not only that, the Plexus joint deflected further than either of the others, under substantially less load.




So to sum up, if I use the Plexus joint as described in the above thesis, what I get is a joint that deflects more, under less load, and is prone to inducing (potentially catastrophic) panel fracture, as compared to traditional fillet and tabbed joints.


But to answer your question, no I didn't read the entire thesis. The mathematics were way over my head (hence possibly redeeming I suppose), but most of the prose was pretty sophomoric, unclear and in-concise.

Kinda what I expected from the examples you posted.

But I did read the T-joint (and other sections) in depth.

So that leads us back to your assertions.

It could be construed that you have some hidden, unfathomable agenda.

It could be that you're a crypto-engineer, i.e. you have to believe the results to see them. (Or in this case misrepresent them)

Or you could (somewhat perversely) enjoy making other people use their brains.

Or any number of other postulates.

But in this, as shown by the thesis you cherry-picked, and the real world experience of many people (for me not against the Plexus but for the bonded, filleted and tabbed), the evidence is overwhelmingly against the idea that Plexus is better in these applications.

But hey, chose for yourself. That's part of the reason I entered this discussion with an example of failed tabbed joints.

Remember, in the thesis, there is a difference between failure of the joint and failure of the bond.

Look at your bolded statement above, consider the "weakness" of the test panels, and then think about the results again in a situation where the panel strength is greatly increased.

The only agenda I have on this point is to question why people say a plexus bond is structurally inferior to a traditionally tabbed bond - when the evidence says otherwise.

In the end, the thesis indicated that a bonded AND tabbed joint is the best method. And I think that's exactly what you see in the industry for the most part. Which means that, no, traditional tabbing is not the only, or best method.