Those of you who have been readers for a while know that I’m a bit of a tool freak. That said, like most cruisers, I also tend to be conservative about stuff that costs money (cruising is expensive enough without being frivolous), takes up space, and adds weight to our boat.
And I like to think that after some five decades of maintaining boats, I’m pretty smart about balancing those conflicting needs well.
But this year I have been engaged in one of the most complex projects I have ever tackled—more details in a future article—during which I upgraded four of my tools.
And after using those new tools I realized that I should have bought them years ago, since they add to the quality of my work and make it go more quickly and easily, too—truly a win, win.
Here they are:
Knipex “pliers wrench”… Cannot imagine life aboard without them. A 5” size pair lives in on a strong magnet in the galley as the world’s greatest pot grips (we use a lot of no handle pots) and is constantly tasked with pliers and wrench like duties throughout the boat – marital harmony required a second pair (actually we have three different sizes in stainless aboard). One of those tools that once you have one, one can’t really imagine life before… Expensive but worth every penny: eg/see https://www.amazon.com/KNIPEX-86-03-250-SBA/dp/B005EXOK22/ref=asc_df_B005EXOK22/?tag=hyprod-20&linkCode=df0&hvadid=309811990469&hvpos=1o2&hvnetw=g&hvrand=16233023170579375023&hvpone=&hvptwo=&hvqmt=&hvdev=m&hvdvcmdl=&hvlocint=&hvlocphy=9057137&hvtargid=pla-491566793747&psc=1
Hi Christopher,
Looks like a nice tool. The only thing I worry about with tools like that is the temptation to use them in places where a properly sized wrench or socket is the right solution. That said, that’s more about self discipline and good tool skills, rather than the tool itself—definitely something I have had to work at over the years.
I whole heartedly agree with Chris. I am a plumber by trade and stumbled upon those Knipex smooth jaw pliers about 6 years ago. They are unique for pliers in that they have a high leverage cam engineered into the design. So when you are squeezing the handles to apply torque your squeeze on the object is multiplied. And you need that since they are smooth jawed with no teeth to bite. Not only are you leaving the work surface undamaged (huge plus!), but the added leverage makes rounding over corners a non issue. Add a hard chrome finish for rust resistance and you have a serious professional tool. In fact, they can almost replace adjustable wrenches. I can’t recommend Knipex enough, especially their “Cobra” line of pump pliers.
P.S. I own the 300mm version of the pair Chris provides a link to and if you can only buy one size, that is the one I recommend. Cheers, Marc
Hi Chris,
Interesting, I had not seen these. A sort of parallel tool that I really like is a locking crescent wrench. It is less likely to round off, you can let go of it, it is great for 1 person jobs and it is easier to take it on and off as you don’t constantly need to adjust.
https://www.amazon.com/Stanley-85-610-10-Inch-MaxGrip-Adjustable/dp/B00009OYGZ/ref=sr_1_8?crid=6PEUN7N211J9&keywords=locking+crescent+wrench&qid=1578332323&sprefix=locking+crescent+%2Caps%2C137&sr=8-8
Eric
I’m also a strong vote for Knipex. I find that often I need 2 wrenches of the same size and there’s only 1 in the set, of course the socket is somewhere far away or in use. I can keep 1 Knipex (personal fav is the 250mm) with the wrench roll-up and cover 95% of all use cases. To John’s point about the right tool at the right time, if you’re not turning with them and only holding your options are wide open.
Torque wrenches -truly indispensible for high performance and therefore highly stressed equipment. I moved over to an electronic one that sits between wrench and socket and was a much, much cheaper and also space saving solution. And before I atacked any crucial tasks I checked the calibration – spot on.
Hi Trevor,
That’s interesting. I did not even know there was such a thing. Could you link to an example.
e.g.
https://smile.amazon.com/ACDelco-ARM602-3-Digital-Adapter-Audible/dp/B004VYUKTC/ref=sr_1_2?crid=3VGX5I8XLFMBK&keywords=3%2F8+torque+adapter&qid=1578679484&sprefix=3%2F8+tourque%2Caps%2C213&sr=8-2
Hi James,
Thanks for the link. Certainly an interesting alternative. I think that I will stick with my wrenches though since I would worry that adding that gadget would make accessing some of the tighter areas on my boat more difficult.
Emphatic yes to 0.01g digital scales; this has made a significant difference to epoxy mixes (that and an eye-dropper for part B). My must-have tool? a “resin disk” to fit a brushless Ryobi 18V grinder—I have a wood boat and this $12 AUD tool, plus cheap 80 grit disks, is the wood-shaper sans pareil.
@trevor allen—can you link to the electronic torque ‘inter-wrench’?
Kit, there are numerous “torque wrench adapters” out there in the wild, just google for this term. They work based on a simple wheatstone bridge which measures the elongation of the wrench extender which is the main part of the adapter.
Note that these adapters give you a more-or-less accurate reading of the force you currently produce, but they will not stop you over-torquing the bolt as they just _show_ the torque, they do not limit it as a “real” torque wrench does.
You might want to look here https://youtu.be/XSsvcNtUGtA for a disassembly and explanation (had a good smile when looking at the video)
Hi Kit,
The eye dropper is a great idea. And I love the Ryobi grinder I just got. Have to look into the resin disk. Could you link to the one you use?
Hi John,
Nice addition to a niche of cruising where light is rarely shown.
Someone correct me if I am wrong as I am well out of my pay grade, but another reason for achieving the correct torque is that the accumulated sheer strength of numerous bolts holding, say a chainplate up against a fiberglass hull, is far less important than the friction generated between the surfaces of the chainplate against the hull: friction that is largely dependent on the torque of the bolts.
This flew against my assumed initial understanding (that the sheer strength of the bolts was the important element), but emphasized to me the import of correct torque.
My best, Dick Stevenson, s/v Alchemy
Hi Dick,
You are right, and like you I was surprised when I learned that. More coming in another article on the rebuild of my JSD chain plates that I just completed.
Dick/John,
You are correct, but it is far more complicated than that and depends on several factors. Realistically, with a chainplate on an FRP or Al hull you will never tension a bolt into it’s elastic limit without damaging the hull, so technically you are still in a ‘snug tight’ state, so the ultimate load capacity will be determined by some combination of friction and shear, though with adequate backing plates on Al you’d be close.
As John alluded to in the article, it’s very hit and miss trying to ascertain bolt tension from torque in the field on used equipment, outside of a controlled factory environment where thread condition, surface coatings (zinc is slippery under load), lubrication, ambient temperature, etc are all known, it’s really very much a guide only. Many bolt heads have popped off long before the supposed required torque has been reached, usually due to over lubrication.
As an aside, if ever taking the head off of your engine, always replace the bolts with new ones as these are factory installed with a controlled ‘torque plus angle’ method that takes them just past the elastic limit into plastic deformation, hence they are ‘overstretched’ once used.
Hi Dan,
Good point. I will be writing more about chainplates in the future and will look forward to hearing your obvious expertise then. (The only reason I dare write about such stuff is that I know we have members with real expertise in these areas who will correct any of my mistakes before damage to others is done.)
Trying to hold a shear load by the shear strength of the bolts is very bad design. In practice, it is very unlikely to get more than one or maybe 2 bolts to load in shear (run out of hole slop) at exactly the same displacement so any other bolts are just decoration.
And you can always design proper stretch if one is determined enough. This can run into things like necked down bolts to get a “soft enough” bolt. Remember a critical rule of bolted joints: the joint needs to be stiffer than the bolt. Therefore o-rings in grooves make for a better bolted joint than cork or paper gaskets. Ad nauseum.
Hi Dan,
In a straight section of an aluminum hull with a proper backing plate it should definitely be possible to clamp hard enough to hold by friction but I share your concern of doing it with FRP. FRP also has the nasty little habit of creeping so it takes some pretty careful design to end up with a sufficiently high tension over time.
To deal with the variability in K factor, I am a huge fan of doing some form of known lubrication and adjusting torque appropriately. When I was doing large industrial machinery, every fastener had a prescribed application of a specific anti-seize. With my current company, at the urging of a few of us, every single fastener gets a weak loctite unless something else like a strong loctite or a grease is required. The advantages are: better known K factor, greatly reduced corrosion concerns for service and less chance of a poorly designed joint having the fastener(s) back out (unfortunately there are way too many examples of this, bolt stretch length seems to be often forgotten).
Eric
Hi Eric,
That’s interesting. What K factor do you use for a light application of loctite? I ask because I’m a total fan of the stuff but was not able to find any good data on the change in torque for loctite.
On attaching a chain plate to GRP, what about attaching a good size G10 backer plate on the inside with epoxy? At least with G10 we know the deform load and it seems to me that this would spread both the crush load on the laminate and distribute the forces over a larger area of the hull. Should bond well to the hull too. Of course all core, if present, would need to be removed and the laminate built back up.
Hi John,
For the level we are talking, probably the easiest thing would be for people to use Loctite 243 as a general purpose. Some people switch to 222 for bolts under 1/4″. When applied per their instructions on a stainless bolt, a K factor of 0.2 is reasonable for calculations. I then combine this with shooting for 75% of yield strength of the bolt. Loctite actually publishes values for their anti-seize compounds and most have a k factor around 0.15 but there is some variability. Certain greases can get it down to about 0.1 but I haven’t ever seen lower than that.
When I am figuring out torque on a fastener, I look at a few things. Realistically, once you do it a few times and learn the rules of thumb, you can skip most of this and only calculate for the close cases.
– Bolt tensile strength (for most types you just look at shank stress but there are some you need to look at popping the head off)
– Bolt head torque capability (hex drive flat heads and pan heads typically are limited here)
– Bolt thread stress
– Mating part thread stress
On attaching to GRP, as you point out with proper design it is possible not to crush the matrix (most designs will always have a bit of local crushing). My concern is maintaining the bolt tension over time. The amount that a bolt elongates when tensioned is actually pretty small so a small amount of creep can take this tension way down over time unless the bolt is really long and the fiberglass really thin. 2 tricks that I have used when confronted with this is to require retensioning as part of a yearly PM program (preventative maintenance) and to put a spring into the system such as the appropriate stack of disc washers under the bolt head but both of these require calculations to make sure they will work. It is not that it isn’t doable but it is trickier to get right.
Eric
Hi Eric,
That’s great, thanks very much. Good tip on bolt tension over time. I just added and upsized the bolts on out JSD chain plates and had to set the torque four times over two days before it stayed at the desired setting. I’m guessing this was because I have a thin layer of BoatLife between the plate and toe rails for isolation. Based on your comment I have just made a note to torque again in the spring.
Another question: The more I learn about this, the more I understand that there is a lot of uncertainty and that only gets worse when laminates are involved. Do you have a feeling for a good safety margin on shear failure in these cases? I have found a good calculator online (see further reading) but obviously it’s only going to be as accurate as the inputs, so a hefty safety margin seems indicated.
Hi John,
Yes, there is a lot of uncertainty. A lot of the work that engineers do is to understand that uncertainty and make sure that in a reasonable worst case scenario, the uncertainty can’t cause a failure. Tolerance analysis is one example of this and structural analysis is another example, you should always run the nominal case but you also need to run other cases as well. Sometimes things are well known, when I did air compressors, the loads were just based on air pressure and there are multiple safeties to limit it whereas if you are designing something like JSD chainplates, the actual loads are not well known and you hope Jordan was conservative enough.
Is your question what we should use for a safety factor in a shear joint when the exact laminate is not known? Assuming I have this right, I generally do not apply a single safety factor to account for everything, I apply known factors as possible and then a much smaller safety factor that deals with the uncertainty.
The way that I work is to first understand my loads (converted into stress) and if fatigue, wear, corrosion or anything else is involved. Then, I understand my material properties and calculate a stress limit that represents the maximum stress you can apply before it would fail with all of the other stuff going on like fatigue. Examples might be if it is steel and in fully reversing fatigue, the stress limit will be ~50% of the yield strength and there may also be allowances for temperature, corrosion, wear, etc. Finally, I apply an appropriate safety factor between the load and my calculated stress limit. If the loads are well known, the material is well defined, and the part(s) are reasonable to analyze, I might run a safety factor as low as 1.5. Other times it might be 3-4. I have parts where my calculated safety factor is 1.5 and the static safety factor if you just look at stress and yield is over 5:1 and this difference is due to the knock down factors I am running on the material strength. Because I work in regulated industries, it is very rare to be able to run a static safety factor below 3 on anything but that is a pretty simplistic way of looking at it.
One rule that I work to with bolted connections is that I try to have the preload on the connection be double the expected separation force but there are times where I have intentionally run much lower numbers. If the force was not well known, then I would increase the preload.
Given the stated problem, I would tend to assume a pretty weak layup and then run a reasonable safety factor like 3 to yield and see where everything ends up. Sometimes you find that you end up with an unreasonable design and then need to start sharpening your pencil and moving past very general calculations and other times it is reasonable and you just build it.
Eric
Hi Eric,
Thanks, that’s exactly what I wanted to know. And well explained. What we don’t know, unfortunately, is what safety factor Don Jordan already applied to his design loads, so we may be piling safety factor on top of safety factor. Still, better that than the other. Anyway, the take away for me is that we need a big safety factor on the laminate sheer and pull through because of the uncertainty on said laminate is much more than on the chain plate or bolts where we can get a lot closer. That really helps.
Ballpark for endurance limit is 50% of ultimate tensile strength (UTS), not 50% of yield though. Good discussion otherwise. 👍
Hi John,
As you can probably guess, I am not a fan of the concept of a design load. For one, I am unaware of a standard definition of it and as such, it is unclear whether it incorporates a safety factor or not. Also, it tells you nothing about whether the the load is cycling or not. For the JSD example, you are mixing several materials and types of loading so a design load which applies 1 blanket static safety factor would yield a design which is quite uneven in terms of how close to failure all the parts are. For example, if nylon line is used, it has extremely poor fatigue performance whereas the stainless bolts for the chainplates are much better so this should be taken into account. While providing a blanket safety factor applied everywhere such as in a design load or in a SWL can make it easier, it does not result in the optimal design. Doing it the way I like to does take more time and also it can be tricky to figure out certain information like there are a lot of materials people use that do not have publicly available S-N curves or if you think that fatigue could be a concern, actually getting numbers for how many cycles at different loads can be tricky.
Eric
Correction: “many” head bolts are plastically deformed (also known as “torque to yield”). This should be specified in the manual and clearly denoted as single-use bolts. This can also apply to main cap bolts, rod cap bolts and other critical fasteners. But it also may not apply so RTFM or maybe buy an unneeded set of new head bolts.
“pliers wrench”… Cannot imagine life aboard without them. “ thanks Christopher. Did not know what they were but after looking them up I agree!