Autonomous
Turgid Member
A structural fillet couldn't hurt. The JSD will introduce dynamic loads the designer and builder never anticipated.
Beefing up the insides for JSD
- Thread starter bilbobaggins
- Start date
bilbobaggins
Member
- 65
- 5
- Thread starter
- #22
I'm quite happy to introduce a 'structural fillet' around the inside corners. That's easily done by wrapping some of the layers of cloth round/into the corners.
Aircraft engineer Don Jordan recommends the optimum 'throat angle' of the bridle arms is about 20 degrees - or 10 degrees either side of the longitudinal, so that's the angle at the transom/hull. The system is INTENDED to pull the stern back into line, until the ambient/transient load is equalised into the two bridle legs.
I've studied in depth all DJ's design notes and comments that survive, all the reams of published user-experience over several decades, and spoken to several experienced users of JSDs. There are also the views of those who make and sell JSDs, some of which - like builders and plumbers - exhibit resistance to innovative and better materials.
There is no need for 'bungee' elastic, stretchy nylon, or similar to absorb energy. The function-mode of the JSD supplies that 'progressive resistance to acceleration' by design. And it works. The only way a sudden shock load could be created would be if a submarine was snagged. I'll live with the statistical probability....
Grateful thanks to all above for valuable input - and especially to 'longy', and 'Zonker' for putting some numbers onto the task, which gives me confidence to do it like that. I'll aim to accommodate a 'Design Load' of 9000lb/40kN.
This is one modification to/for my boat where 'failure is not an option'.
Aircraft engineer Don Jordan recommends the optimum 'throat angle' of the bridle arms is about 20 degrees - or 10 degrees either side of the longitudinal, so that's the angle at the transom/hull. The system is INTENDED to pull the stern back into line, until the ambient/transient load is equalised into the two bridle legs.
I've studied in depth all DJ's design notes and comments that survive, all the reams of published user-experience over several decades, and spoken to several experienced users of JSDs. There are also the views of those who make and sell JSDs, some of which - like builders and plumbers - exhibit resistance to innovative and better materials.
There is no need for 'bungee' elastic, stretchy nylon, or similar to absorb energy. The function-mode of the JSD supplies that 'progressive resistance to acceleration' by design. And it works. The only way a sudden shock load could be created would be if a submarine was snagged. I'll live with the statistical probability....
Grateful thanks to all above for valuable input - and especially to 'longy', and 'Zonker' for putting some numbers onto the task, which gives me confidence to do it like that. I'll aim to accommodate a 'Design Load' of 9000lb/40kN.
This is one modification to/for my boat where 'failure is not an option'.
Last edited by a moderator:
I've installed two JSD anchor systems, so I can appreciate what you're going thru. You don't want failure. But short of having the build sced for the stern laminates and the a qualified laminate engineer, how do you satisfy (and pacify) the boat owner? Neither of the install's I did ever got used, one at least got tested under moderate conditions.
The corner of the hull transom intersection and the curve of that joint add a lot of strength to the transom area. The transom panel is curved aft - so any compression bending of the transom would be the transom bowing further aft. The backstay chainplates are just around the corner from the JSD padeye, so I suspect that area AND the corner are already built up quite a bit. A careful inspection of the inside will help ascertain how much has been added by comparing to the joint about a foot down. The backstay chainplate bolts will be quite close to the edge of the corner fillet, so to add lams to the corner will require removing bolts to allow the layer to extend out onto the transom far enuff to add strength.
The corner of the hull transom intersection and the curve of that joint add a lot of strength to the transom area. The transom panel is curved aft - so any compression bending of the transom would be the transom bowing further aft. The backstay chainplates are just around the corner from the JSD padeye, so I suspect that area AND the corner are already built up quite a bit. A careful inspection of the inside will help ascertain how much has been added by comparing to the joint about a foot down. The backstay chainplate bolts will be quite close to the edge of the corner fillet, so to add lams to the corner will require removing bolts to allow the layer to extend out onto the transom far enuff to add strength.
bilbobaggins
Member
- 65
- 5
- Thread starter
- #25
All this is much appreciated, people.
I misspent much of my innocent youth with my tender pink little bottom strapped to an ejector seat. I didn't ever need to deploy the thing, despite coming darn'd close on occasion. "When all else fails, put your trust in God and Martin-Baker"
I view the JSD in much the same light.
I misspent much of my innocent youth with my tender pink little bottom strapped to an ejector seat. I didn't ever need to deploy the thing, despite coming darn'd close on occasion. "When all else fails, put your trust in God and Martin-Baker"
I view the JSD in much the same light.
CaptainAhab
Anarchist
There is a big issue that hasn't been mentioned.
You've got five options. Don't do it, Do it wrong. Almost do it right, Do it right, Over do it in terms of installation.
The last two might save the day. The first three may sink the boat. Unless you get your attachment point right on the shear line like in that picture, you are risking water coming into the hull if your attachment point fails.
You've got five options. Don't do it, Do it wrong. Almost do it right, Do it right, Over do it in terms of installation.
The last two might save the day. The first three may sink the boat. Unless you get your attachment point right on the shear line like in that picture, you are risking water coming into the hull if your attachment point fails.
phill_nz
Super Anarchist
bilbobaggins
Member
- 65
- 5
- Thread starter
- #29
You wouldn't believe the internet and YouTube 'rabbit holes' I've been down, looking for illumination, resolution and perhaps absolution.
Then I found this....
Then I found this....
bilbobaggins
Member
- 65
- 5
- Thread starter
- #30
I'll explore that. It would be easier to do now, as the boat is 'high and dry'.
An insurance surveyor will want to poke about in there, anyway, and if I can show evidence of 'enhancement' s/he probably won't poke any further.
Last edited by a moderator:
robtoujours
Member
Relevant:
Shows chainplates at 24:30
Shows chainplates at 24:30
Last edited by a moderator:
bilbobaggins
Member
- 65
- 5
- Thread starter
- #32
That's a pleasing video by James/'atomvoyager' which illustrates several facets with clarity. His boat is of similar size/displacement to mine.
I found myself stopping/rewinding at several points, which I'll offer up for comment.
He chose to use 5/16 or ~8mm dyneema. I purchased 10mm, not least due to its easier 'manual handling'. The cost difference was, to me, insignificant.
I chose a continuous 'rode + leader' with no join. I can't see the point of introducing a superfluous join which adds nothing and may have potential for a splice failure.
I, too, made my bridle from the same diameter rope as the rode. Why not?
The Titan shackles used are rated at 1 1/2 tons. The rope rode - several times that.
I used tubular s/s shackles so the spliced dyneema loops are captive under load.
Every rope manufacturer I consulted - New England, Marlow, Southern, English Braids, and John Franta of Colligo Marine- advises that joining ropes - rode to bridle - using knots e.g. cinch hitches is a 'no-no'. That is throwing away up to 50% of line strength. Yes, it's neat and quick, but....
James shows the now-accepted means of providing a 'lock point' for attaching a recovery line via a rolling hitch. Simple, cheap, effective.
James shows one method of deployment.... end-weight first. Some experienced users advocate feeding out the leader and 'closest' cones first, with end-weight last.
James shows and uses excellent purchased 'beefed-up' cones. Earlier DIY lightweight examples frayed badly due to violent 'flutter' and needed replacement.
The neat chainplate arrangement used has but 3 through-hull bolts. Sufficient? If so, how...?
FWIW.
:huh:
I found myself stopping/rewinding at several points, which I'll offer up for comment.
He chose to use 5/16 or ~8mm dyneema. I purchased 10mm, not least due to its easier 'manual handling'. The cost difference was, to me, insignificant.
I chose a continuous 'rode + leader' with no join. I can't see the point of introducing a superfluous join which adds nothing and may have potential for a splice failure.
I, too, made my bridle from the same diameter rope as the rode. Why not?
The Titan shackles used are rated at 1 1/2 tons. The rope rode - several times that.
I used tubular s/s shackles so the spliced dyneema loops are captive under load.
Every rope manufacturer I consulted - New England, Marlow, Southern, English Braids, and John Franta of Colligo Marine- advises that joining ropes - rode to bridle - using knots e.g. cinch hitches is a 'no-no'. That is throwing away up to 50% of line strength. Yes, it's neat and quick, but....
James shows the now-accepted means of providing a 'lock point' for attaching a recovery line via a rolling hitch. Simple, cheap, effective.
James shows one method of deployment.... end-weight first. Some experienced users advocate feeding out the leader and 'closest' cones first, with end-weight last.
James shows and uses excellent purchased 'beefed-up' cones. Earlier DIY lightweight examples frayed badly due to violent 'flutter' and needed replacement.
The neat chainplate arrangement used has but 3 through-hull bolts. Sufficient? If so, how...?
FWIW.
:huh:
Yes, I noticed that. Have to check on sheer strength of bolts AND know how thick the laminate is around them. I have seen race boat deck hardware that was migrating along a deck by ripping thru the laminate - not a pretty sight. And are the bolts threaded where they go thru the lam or are they full dia shank?
I think he said 3 x 3/8" bolts. (9.5mm)
Let's assume he bought 316 and they were full shank through the laminate.
Cross sectional area for 3 bolts = 3 x pi (9.5)^2 / 4 = 213 mm2
Tensile is about 579 MPa. Shear strength is about 60% = 347 MPa
Failure load = 347 MPa x 213 mm2 = 73911 N =7534 kg = 16,575 lbs.
If we figure peak load ~ displacement of the boat he's got an ~2 safety factor with new bolts. (I assume the load is not necessarily split by the bridle because an off axis load could really load up one side and the other leg goes close to slack.)
Notice I said "new bolts". Crevice corrosion etc and your bolt after 5 years of wave splashes could be not as strong as new.
If it was my boat I would probably want at least 1 more bolt - who knows how they are load sharing? If they were clamping metal parts together that are relatively stiff compared to the bolts you would have good load sharing but in a thick glass laminate I'm not so sure that the first bolt doesn't take a slightly bigger load.
Let's assume he bought 316 and they were full shank through the laminate.
Cross sectional area for 3 bolts = 3 x pi (9.5)^2 / 4 = 213 mm2
Tensile is about 579 MPa. Shear strength is about 60% = 347 MPa
Failure load = 347 MPa x 213 mm2 = 73911 N =7534 kg = 16,575 lbs.
If we figure peak load ~ displacement of the boat he's got an ~2 safety factor with new bolts. (I assume the load is not necessarily split by the bridle because an off axis load could really load up one side and the other leg goes close to slack.)
Notice I said "new bolts". Crevice corrosion etc and your bolt after 5 years of wave splashes could be not as strong as new.
If it was my boat I would probably want at least 1 more bolt - who knows how they are load sharing? If they were clamping metal parts together that are relatively stiff compared to the bolts you would have good load sharing but in a thick glass laminate I'm not so sure that the first bolt doesn't take a slightly bigger load.
estarzinger
Super Anarchist
- 7,505
- 898
just a thought - did you give any consideration to building essentially aft facing composite chainplates. Composite chainplates are typical and normal today on race boats (it is also common to have composite emergency rudder gudgeons). You are talking about enough composite work already beefing the area up, why not just do it 'the modern way'?
We had aft facing chainplate integral in Hawk's toerail and tied it to the frames for the jsd - along with 4 lifting strong points further forward in case we ever needed to be craned out.
As an aside - I personally am not a huge fan of having to hang a bit over the side or stern of the boat in breaking wave conditions to screw two shackles on.
..............................................................................
I'm just holding my mouth shut about dyneema issues - some of it makes me want to cry but i'm not Sisyphus, I stopped trying to roll the boulder.
Last edited by a moderator:
bilbobaggins
Member
- 65
- 5
- Thread starter
- #37
I'm grateful again to 'Zonker'. Substituting for my 4 BUMAX88 bolts, I note ~18608kg/40937lbf, which should be 'quite sufficient'. I hope....
"Minimum specification for BUMAX 88 is a tensile strength of min. 800 MPa and a yield strength of min. 640 MPa. However, in reality BUMAX 88 counter-sunk bolts seem to be (on average), closer to around 960 MPa (tensile) and 860 MPa (yield)."
Among the many 'rabbit holes' I've dived down, searching for education and illumination, is this one: 'Simplified Procedure for Designing Bolted Composite Joints - NASA Technical Memorandum 100281' ( https://ntrs.nasa.gov/api/citations/19880005638/downloads/19880005638.pdf
At my time of life, this 'bear of little brain' shouldn't be found delving into such deep matters ( 'The Tao of Pooh' - A A Milne )
Nevertheless, the notion of fabricated composite chainplates had escaped me - and it shouldn't have, for I raced a RORC Series on 'Triohe', a hand-built carbon Farrier F9AXR, which had plenty of those. Thanks again to Evans S. for the fresh heads-up. I have further uses and places for that idea.
People will be wondering how on earth I came to be using Harken Specials on my little old MAB from last century, and that's a tale best suited to a starry night at anchor, a bottle of uisge baugh, and some crystal glasses. Suffice to say that Old John Foulkes, late of The Chandlery Barge, Hamble is implicated.
I plead that I have the things. Here below is part of the haul - quite what I'm going to do with the thing in the middle is a conundrum.
"Minimum specification for BUMAX 88 is a tensile strength of min. 800 MPa and a yield strength of min. 640 MPa. However, in reality BUMAX 88 counter-sunk bolts seem to be (on average), closer to around 960 MPa (tensile) and 860 MPa (yield)."
Among the many 'rabbit holes' I've dived down, searching for education and illumination, is this one: 'Simplified Procedure for Designing Bolted Composite Joints - NASA Technical Memorandum 100281' ( https://ntrs.nasa.gov/api/citations/19880005638/downloads/19880005638.pdf
At my time of life, this 'bear of little brain' shouldn't be found delving into such deep matters ( 'The Tao of Pooh' - A A Milne )
Nevertheless, the notion of fabricated composite chainplates had escaped me - and it shouldn't have, for I raced a RORC Series on 'Triohe', a hand-built carbon Farrier F9AXR, which had plenty of those. Thanks again to Evans S. for the fresh heads-up. I have further uses and places for that idea.
People will be wondering how on earth I came to be using Harken Specials on my little old MAB from last century, and that's a tale best suited to a starry night at anchor, a bottle of uisge baugh, and some crystal glasses. Suffice to say that Old John Foulkes, late of The Chandlery Barge, Hamble is implicated.
I plead that I have the things. Here below is part of the haul - quite what I'm going to do with the thing in the middle is a conundrum.
This is a Harken hi load padeye. The fasteners are 10mm. The pitch distance between the 2 fasteners in line with the load is 33mm. This is not enough.
The general rule of thumb for bolted fittings in composite is 4x-5x fastener diameter spacing.
When they get closer together (3.33x) then the bolts start tearing out instead of the laminate failing in bearing stress. Now most boats probably don't load up the padeye close to the breaking strength loads, but it's why bolts in chainplates are widely spaced.
Composite chainplates are great (and easy to calculate) if the load is all in one direction, like, uh a chainplate. You always know the load direction because the rigging loads are a constant direction.
I'm not sure how good they are with off-axis loading, like a JSD. The load can be vertical say +/- 20 degrees (?) or angle inboard or outboard depending on how the boat is slewing down a wave. The unis in a composite chainplate are very strong along their long'l axis, but strength drops off very fast when not in line.
Here is what I am talking about (this is a well thought out chainplate). The load can go inboard/outboard (red) or up/down (green) quite a bit.
For a JSD I think a metal plate makes a lot of sense. Most people's composite skills are not good enough to make a really good uni carbon or glass chainplate. They are good enough to add a lot of glass on the inside of the hull and drill some holes in the hull.
The general rule of thumb for bolted fittings in composite is 4x-5x fastener diameter spacing.
When they get closer together (3.33x) then the bolts start tearing out instead of the laminate failing in bearing stress. Now most boats probably don't load up the padeye close to the breaking strength loads, but it's why bolts in chainplates are widely spaced.
Composite chainplates are great (and easy to calculate) if the load is all in one direction, like, uh a chainplate. You always know the load direction because the rigging loads are a constant direction.
I'm not sure how good they are with off-axis loading, like a JSD. The load can be vertical say +/- 20 degrees (?) or angle inboard or outboard depending on how the boat is slewing down a wave. The unis in a composite chainplate are very strong along their long'l axis, but strength drops off very fast when not in line.
Here is what I am talking about (this is a well thought out chainplate). The load can go inboard/outboard (red) or up/down (green) quite a bit.
For a JSD I think a metal plate makes a lot of sense. Most people's composite skills are not good enough to make a really good uni carbon or glass chainplate. They are good enough to add a lot of glass on the inside of the hull and drill some holes in the hull.
Autonomous
Turgid Member
Which is why structural fillets on the inside corners of the transom-hullside corners are worth considering.