When a bracing through rod is not perpendicular to the trunk surface the washer will bear on a wood surface on only part of its surface. The old practice was to counter sink a flat bearing surface into the wood. ANSI-2000 A300 Part 3 now require countersinking into bark only. That may leave the washer without a flat bearing surface, particular in larger diameters.

Has anyone used beveled washers, like those used in steel I-beam construction to provide a flat bearing surface? Sources? Specifications?

Reply to post by Scott Cullen, on July 27, 2000 at 14:10:01:

Hi Scott, there was a recent discussion of this topic in "The Journal of Arboriculture",Vol.25, no's 2,4 and six. Using beveled washers or pieces of tubing cut at an angle will not negate the effect of the rod pulling laterally through a branch ( along the grain, that is). The only way to stop that is to countersink so the washer bears at 90 degrees to the rod and is flat against the bearing surface of the branch. James.

Reply to post by James Causton, on July 27, 2000 at 14:10:01:

I'll have to dig up those issues of JoA. Sounds like a dilemma. If hardware not 90 degrees MUST have a countersunk flat to be sound and A300 says we MUST NOT coutersink into wood and A300 often requires through hardware (e.g. cable anchors) to be other than 90 degrees to wood what are the alternatives?

I would have thought the hardened, beveled washer (not so sure about cut tubing which might crush) would have the effect of transmitting the force to the wood as if it was countersunk... the washer is bearing on a flat surface . Can you explain a little further?

Reply to post by Scott, on August 06, 2000 at 18:52:58:

OK.let me try and expain it more clearly.
If the through bolt passes through the limb at any angle other than 90degs. there will be a natural tendency for the rod to pull along the grain as it is tightened (shortened). I do not know at what point the torque on the nut makes this critical, it may be at a point well above the needs for a cable brace.

Reply to post by Scott, on August 06, 2000 at 18:52:58:

The twisting effect is the result of the vector forces. The top installation shows the forces of the washer and the wood pressing against each other. They are directly opposite, and cancel each other (assuming the wood fibers don't crush or buckle from too much pressure).

The second installation is at an angle with bevelled washers. The force on the washer is parallel to the rod, and at an angle to the wood surface. The wood presses back perpendicular to the seat of the washer- not parallel to the rod. The result is a directed vector toward the long side of the bevelled washer. This forces the rod to twist within the pilot hole, and increases pressure against the sides of the hole.

Whether this actually represents a problem will depend on the forces actually applied, the ability of the wood to resiste splitting, etc.

Reply to post by Russ Carlson, on August 07, 2000 at 12:27:53:

I understand the forces James described now.

1. You've exagerated the oversize nature of the clearance hole to make it apparent in the graphic. A-300 2000 allows no more than 1/8" over (1/16 each side), but 1/16" over (1/32" each side) is traditional practice. In my experience in actual practice large diameter rods over any but the shortest distances don't really float in the clearance hole, you usually have to turn them through with the drill. I'd suggest the splitting force is largely disapated in sidewall friction and rod distortion.

2. The bigger question for me is what happens to the forces withou the beveled washer. I you can link the graphic for me the following observations apply. Only one point of the washer is taking the force from the nut. The force at the green arrows will either be concentrated toward the tree and have a crushing effect or be disapted attempting to bend the washer or at least applying friction to the nut (torque goes into crushing or resisting friction rather than movement along thread). There will be a tendancy to twist the nut unevenly against the threads (red arrows) rather than moving the nut along the threads.

If part of the objective is to close or restrict an already drawn up (cables above) crack (shown exagerated) the desired effect is closing the space between nut & washer pairs, i.e. moving them along the threads. Both the movement and the closure require force and it cannot be productive to lose that force in extra friction (thread and washer) and in crushing wood at one side.

3. Intuitively, the beveled washer provides a flat bearing area for the washer and solves more problems than it creates.

Scott

Reply to post by Scott, on August 15, 2000 at 01:31:02:

With or without a beveled washer/spacer ends up being a trade off in respect of what gets damaged. Using the washer, at some point of torque on the nut, the wood gets damaged. Without the washer, at some point the bolt/threads/nut gets damaged. Do we need to apply that much torque?? Probably not, but the tensional loads which the installation could be subject to in a very strong wind could well do it. James

Reply to post by Scott, on August 15, 2000 at 01:31:02:

1) Correct- the hole is exaggerated. The splitting force is still there, however, and as Mattheck points out (and Wessolly concurs) the forces are greatest at the points farthest from the pivot, or at the outer edge of the stem. Perhaps the forces are dissipated over the entire length of the rod that is incontact with wood, but the force of the rod and the resistance of the wood are still the same, just applied differently. It then depends on how the tree distributes the forces- a factor of the physical characteristics of the wood.

2) The force of the tree (green curved arrow) is directly outward initially, then along the lines of force (parallel to the rod) as the washer crushes inward or the tree grows beneath the washer. The force of the washer is perpendicular to the washer, mostly parallel to the rod except for the minor twisting of the washer. This will yield the same resultant force as the bevelled washer, except that the entire force is concentrated at the point of contact. This means the force from the washer could more easily exceed the force needed to crush the wood fibers, causing damage to the bark and wood at that point.

As the tree grows new tissue beneath the washer (in the gap between washer and stem) the forces will become distributed over a greater area, and the outward force of the wood will be against the washer, not perpendicular to the stem.

This configuration still has the same splitting force as the bevelled washer system when installed. It seems the forces may be dissipated over time by new growth pushing the washer in the opposite direction. The force of the new growth filling the gap may tighten the rod, forcing the two halves of a split stem together. These forces are probably not highly significant in the overall scheme of the system and the tree.

The third installatin possibility is countersinking the washer. This means cutting the wood so it presents a flat bearing surface perpendicular to the rod. Then the outward force of the wood will be directly opposite the washer, balancing it to zero. There will be no twisting of the rod or splitting forces. However, it creates significant damage to the wood, exposing it to potential decay.

3) The bevelled washer and one with no bevel present similar results, except for the amount of initial bearing surface. The bevelled assembly offers less initial damage from crushing. The non-bevelled washer assembly may crush the wood at the point of contact, but may offer slightly more compression as the wood fills the gap beneath the washer. The countersink method is definitely too harmful to be considered, in my opinion.

One last thing to consider- remove the bark or not? I do not think removal of the bark is necessary, and that it is actually harmful. If you look at a split crotch that had included bark, there is often little or no decay in the crotch. Closed wounds show the same effect. The bark becomes included, the tissues eventuall crush and die, and the surrounding wood covers it. Leaving the bark beneath the washer leaves the barrier to decay intact, and the tree can then resist better. you still have the hole for the rod, but it is smaller than the large injury caused by removing bark under the whole washer.

Reply to post by James Causton, on August 15, 2000 at 07:17:06:

Russ and James,

Thanks for follow up. I think the question of applied torque is the key. If the rod is simply slipped through and the nut/washer installed hand tight if that brings them up to bark on both sides this is probably a non-issue (except perhaps under wind load as James suggests). (Then you need to double nut... which the new A300 requires anyway I think... or peen threads to prevent back-off).

If we're talking about 1/2" eyebolts it's probably no big deal... and yes the wood should fill in quickly under the no contact side of that washer.

The real issue comes with large rods, say 3/4 or 7/8. Those washers are probably around 3" in diameter. It's going to take a while to grow into that space. That time period doesn't really matter in installation though and that's the biggest question. IF you have to install with torque to close up a crack. I think that's where there is a trade off and the beveled washer is the better option.

Russ, option 3, counter-sinking, is what we used to do. In the old old days out came the cavity bag and chisels and gouges. In the less old days we made up a special counter-sinking bit. Then we learned about decay and compartmentalization and it might be better not to counter sink unless the near term installation was more important than long term decay. Now A-300 2000 says NO countersinking ever... and that was the presumption at the start of the thread.

I'm not sure I have an opinion on countersinking into bark. What has occured to me though is that if bark is left in place the initial torque will tend to dig a beveled washer into the bark and at least somewhat fix its position relative to the hole... more so with a beveled shim cut from pipe or tube. That might limit the sideways, splitting action of the rod as more torque is applied. I suppose it will somewhat depend on the difference betweed ID of washer and OD of rod.

One last item. When I searched for mamufacturers of beveled washers I learned that they are required by code and standards in steel construction where a bracing rod passes through a beam at an angle other than 90 degrees. I'm unclear if that is A) to balance torque on the threads and make draw up easier and avoid stress on rod or B) to avoid unbalanced point load on the web of the beam.

Scott

Reply to post by Scott, on August 15, 2000 at 01:31:02:

I'm trying to locate examples of these installations. Here's the first one.

Go down to the section entitled "The Details" and look at a cable railing termination in a wooden post. A beveled washer is specified.

Do we think this is purely cosmetic or there is an engineered reason?

As I find more examples I'll post them. I think where I've seen beveled washers before is in timber truss bridges and in timber roof trusses as in a church (that's where the term "cathedral ceiling" came from). So far no good examples to post.

Reply to post by Scott, on August 15, 2000 at 18:17:31:

Scott asked, "One last item. When I searched for mamufacturers of beveled washers I learned that they are required by code and standards in steel construction where a bracing rod passes through a beam at an angle other than 90 degrees. I'm unclear if that is A) to balance torque on the threads and make draw up easier and avoid stress on rod or B) to avoid unbalanced point load on the web of the beam." The answer is that on the angled web of both steel joists and angle iron, the beveled washers are there to keep the compressive forces of the nut (torque) at 90 degs. to the bolt. In steel, the forces exerted on bolts creates a greater friction loading against sideways movement than in wood (shear effect). Plus, in steel, the difference in densities (compressive strength) is very similar so the shear effect of the bolt against the "I" beam web is comparatively reduced.
Please see new approach!!! James.

Reply to post by Scott Cullen, on July 27, 2000 at 14:10:01:

If we are working with cable bracing we have an obligation to understand what is happening with the installions we are using.
If there was an advanced course available in "eyebolt installation" which only cost $12 and lasted 2 hrs. would you take it????
If the answer is yes, then please take a couple of chunks of 4-6in wood. Drill holes through for 3/8 rod, put the rod through and crank them nuts down, observe what happens. Put beveled washers/spacers on and see what happens when you crank them down. As Alex Shigo (aka Uncle Al) said "touch trees".

James

Reply to post by James Causton, on July 27, 2000 at 14:10:01:

I'll try it today and report back.

I do want to repeat my concern is with bracing bolts rather than eyebolt attachments for cables. The compressive forces are higher, the washer diameter is much larger, that larger diameter means the small area of contact is a much smaller percentage of overall washer area, and the force-surface relation ship is changing... % bearing area decreasing as compressive force is increasing.

I completely agree, we must understand what's going on in all of these installations. I appreciate you engagement in this discussion. (You too Russ).

Scott

Reply to post by James Causton, on July 27, 2000 at 14:10:01:

I haven't forgotten. Things just got cluttered this week and no suitable log sections at hand, though I did pick up some beveled washers. I'll get to it and report.

Reply to post by Scott, on August 16, 2000 at 23:05:24:

Promises,promises

Reply to post by Scott Cullen, on July 27, 2000 at 14:10:01:

I promise (really) I'll try your empirical "new approach" James. But in the mean time I've done some research and happened to have an engineer bring up the issue today and asked for his opinion.

The case at hand is the tree to ground guying of a large tree. The engineer calculated horizontal load at 30 feet is 45,000 lbs. That is to be spread among 6 guy cables each going to ground at 45 degrees. ANSI A300 Part 3 2000 requires a bolt to be in line with the guy. The engineer completely agrees. So here we have our topic: Significant load (constant tension on cable or wind on tree creating tension) and Bolt not 90 degrees to tree. It is standard practice to have nut bear on a surface more or less 90 degrees to bolt. This is to apply even tension on the bolt.

In background research I've learned that a bolt in tension is like a spring... it stretches. If you are compressing members together in a joint (in our case closing a split) you'd design the bolt to compress the joint "snug" before there was sufficient torque to tension stretch the bolt. Whether you're compressing two members or just passing throough one you'd also design to get stretch in the bolt before deforming or crushing the member. The stretch is what creates the locking of the not on the bolt. If you deform the member (or particularly if you crush wood fibers that become subject to decay) the initial tension might be lost and the nut could loosen. That's not such a problem if you have a locknut per A300. But if there is a tension load you want it to be even.

Interestingly Shigo in New Tree Biology and Modern Arboriculture explicityly suggests creating a flat bearing surface but minimizing damage to wood. A300 2000 has gone on to say NO WOOD COUNTERSINKING but has ignored the flat bearing surface.

US Coast Guard Standards for the inspection of wooden vessels specifically say that where tension bolts pass through wood members at an angle 90 degrees a flat bearing surface shall be created by countersinking becuase of joint failures that occur where wood fibers are crushed by unequal bearing compression.

So, put all the pieces together. It's not good to crush wood fibers by unequal compresion. It's not good to put unequal tension on a bolt. It's good to have a flat bearing surface. It's not good to create a flat bearing surface in living trees by countersinking into wood. What are the options?

If it's a cable with a small diameter bolt (say 1/2") maybe nothing is needed... the uneven tension is not that great. If it's a larger bolt and the angle is more than minimal maybe we have to ask if the system will be in tension.
If it's a prentive brace bolt that does not have to close up a split maybe you get it "snug" and don't apply more tension and just put on a lock nut. It's just ugly hanging out there with uneven bearing. But then if there's no tension there's also no compression for a beveled washer to apply a splitting or shearing force so it's not an issue anyway.

So that leaves us with the case where there is tension... a high load cable or guy or a brace bolt closing a split. Engineering tells us it's bad practice to unevenly tension a bolt. We don't want to crush wood fibers either. So that leaves the question of whether the compresion against the beveled washers will cause a splitting action before the bolt is tensioned to system load.

Now I'm going to conjecture. First it must be related to the relative diameters of stem and through holes and also the length of sound wood above and below the hole. Drive a nail in a short piece of wood and it will split... shear along the grain. Drive the same nail in a long piece and there may be localized shearing but the crack won't propagate. Also if we're talking about a downward pull along or with grain fibers - as in a guy - we have to consider the splitting potential of a round rod. Picture welding that round rod to a splitting maul and tring to split some firewood.

Second we have to consider additional lateral forces. According to A300 a cable is supposed to be inline with the bolt and shouldn't apply lateral forces. If the installation is substandard in this regard those lateral forces alone could have a splitting effect and beveled washers could indeed compound that. The need for a beveled washer would come from either a vertical angle or a hole drilled off center. To some extent those are controlable instalation variables.

Third, the specifier and installer need to consider priorities. What is the risk of failure if the system is not installed? If the risk is real mitigating it is the first priority and that includes design performance of hardware. A possible or theoretical risk of splitting (or even of decay) becomes secondary though certainly reason for periodic inspection. The liklihood of these secondary risks probably involeves the first point.


OK I'm going to post this. If it requires some clarifying on re-read I'll post a follow up.

Scott

Reply to post by Scott, on July 27, 2000 at 14:10:01:

A good summary of the various aspects to be considered. We seem to be beginning to bring real physics into our thinking and our work.

The last paragraph is the real key, and I think the heart of the A300 standards. It is the matter or priorities. This was discussed on other threads about topping and various other treatments. Which is worse- cutting down a tree, leaving it to fail, or applying a treatment we know may cause some harm, but is less than the harm caused by the alternatives.

To throw one more variable into the equation, let's not forget the lateral forces caused by twisting of the canopy or stem against the hardware. Now we have tension of the bolt (compression of the wood) being pulled in direct line to the cable forces, vertical stress of the bolt not being in direct line, and lateral forces from poor alignment and twisting of the canopy.

You mentioned the effects of splitting, and appied an anology of a nail in a board. One other faccet to consider is the species of tree and integrity of the wood. Somespeceis tend to split easily, others only with great difficulty. Maul swingers learn this quickly. A 'splitting index' would be helpful. Probably already is something like this.

Reply to post by Russ Carlson, on August 28, 2000 at 20:00:01:

Good additional point about canopy twisting in wind Russ. I guess that adds height of attachment to the analysis. The lower you go and the thicker the stems become I'd assume the less twisting there is.

In the system we're designing right now we may be below A300 recommended attachment height but we are well down into non-twisting territory at most attachment points and given ground anchor location and slope we are achieving optimum tree-guy angle inherent in other A300 provisions as well as engineering analysis.

So your point is well taken... the standards have to be interpreted in the context of priorities (A300 Objectives) and all engineering aspects.

Reply to post by Scott, on August 28, 2000 at 23:33:48:

Thinking a little more about your additional, lateral movement point I think we need to recognize both twisting within a member and sway of members relative to each other.

In a flexible (i.e. cable) system energy from TWISTING of members is likely to be lost in cable flex rather than translated to a splitting force. In a rigid (i.e. rod bracing) system twist is likely to apply a splitting force. Particularly in light of recent literature (Smiley I think) recommending brace rods be move above the crotch.

SWAY poses another problem. Even if bolts are properly installed in line with the cable if thw cabled members sway opposite one another there is likely to be lateral force on the bolts, perhaps applying a splitting force to the stems.

Reply to post by Scott, on August 29, 2000 at 07:46:52:

Smiley's work with bolts, and the recommendation to place them above the crotch is basic physics. You have a lever/fulcrum system, with the pivot point at the bottom of the crack (this point can move if the crack continues to propogate). The farther the attachment is from the pivot, the better it will hold (the less force needed to reatard movement). This is the premise for installing a cable EVERY time you install a bolt. And it is the premise for placing the bolt above the crotch- once secured properly, the top of the crotch becomes the pivot point.

Bolts may still be needed below the crotch, too, to limit shear forces and lateral movement caused by twisting of the codominants in the wind.

Reply to post by Russ Carlson, on August 29, 2000 at 12:20:45:

I understand your points which are well made.

The point I was making, following up on James's concerns about splitting is that TWISTING of members could apply a splitting force in rigid (rod) bracing and that the tendancy might be greater with rods above the crotch. Your earlier point is really the key: understand the objectives, the effects and the tradeoffs, not just theory.

Your other comment about cables above rods EVERY time is timely, because I've just had to critically examine this. A300 (Part 3)-2000 provides for it (I'll have to look to see if it's a should or shall). BUT you and A300 both fail to recognize the range of scope of applications. The A300 figure for bracing types shows four bolting patterns all associated with bolts in tension, used to compress a shared crotch-crack between two leaders. Cables above are appropriate in these types of applications.

A300 does not address bolts in compression, or "push" braces, used to keep stems apart (stems which may not have a shared crotch) or used to limit movement of one stem by relying (pushing) on another. A cable above may do nothing to achieve the objectives of the brace. The cable would prevent the stems from moving apart but that's not the objective. The cable would simply go slack under loads moving the stems together.

A300 does not seem to address bolts used to stabilize or "stitch" cracks in single stems or what used to be called "lip" bolts around the mouth of a cavity. While these applications may be passe or arcane, some practitioners or specifiers may interpret A300 and the facts to find them useful. Cables above may not be appropriate (or even feasible) in these applications.

It seems to be a bigger deal in Europe than here, but there are situations that for lack of another term I'll describe as "heritage stumps" where huge old trunk systems with a few remaining laterals of a tuft of sprouts are preserved for heritage reasons. They might require crotch bolting or crack bolting but not even have a structure above to be cabled.

My point is that "EVERY bolt should have a cable above" is an overstatement. That would characterize the types of applications I've described as substandard when in fact and practice the cable above does nothing to achieve the objective or may be impossible to install.

Reply to post by Scott, on September 07, 2000 at 21:06:11:

In paragraph 4 read "a cable would keep the stems from moving FARTHER apart..."

Reply to post by Scott, on September 07, 2000 at 21:06:11:

I stand corrected. Once again I have fallen victim to the human tendency to ASSUME, and accept the narrower view.

EVERY time is an overstatement. I think the vast majority of cases where bolts are used (in US) involve preventing separation of crotches, hence a cable should be installed above whenever possible (as allowed by configuration of the tree structure).

Bolting to maintain a minimum separation between limbs will require different hardware configuration anyway. Placing washers and nuts at the ends of the bolt will do nothing to achieve the goals in this case. The washers and nuts must be placed between the stems to maintain separation. This should be separately addressed in A300, as should the lip bolts or stitch bolts in cavities.

A300 is a good product, as far as it goes. In most cases so far, I think the committee has stopped short of addressing the rare or unusual cases. Language could be crafted that will guide the installer in determining the nature of the system, without actually setting a standard for how it "must" be done. Or perhaps this needs a separate book, as ISA did with the Tree Pruning Guidelines book (1995).

Reply to post by Russ Carlson, on September 08, 2000 at 06:30:36:

Just about everybody agrees that in the typical situation of keeping a crotch from failing if you use rigid bracing rod at the crotch you should also use a cable above (A300 Part 3 Sec. 40.3.1, though it's a should not a shall).

It's certainly belt and suspenders. Maybe we all need to re-examine assumptions. It is intuitive that the leverage of a high cable effectively limits load on the crotch. But let's start with the objective. Typically it is to prevent shear failure of the wood fibers below the crotch if they are put in tension. The objective is not to prevent fracture failure of the leaders above the crotch. The newer literature and the hype about the non-steel cabling systems suggest that it is better not to prevent all tension on and movement of the crotch.... the movement and tension promote wood formation. Our typical approach in supporting crotches is cable first, rigid bracing at crotch as something additional. In that implimentation sequence the cable, steel or otherwise must perform the work if there is no brace at crotch. But if the rigid brace is installed alone will it sufficiently prevent the shear failure below the crotch? Maybe. Remember we're not typically trying to prevent fracture failure above the crotch. What would the cable do? It would limit tension load on the bolt. It would limit compression or crushing load on the wood between the nuts and washers on the rigid brace. Certainly a good idea if the cable has no conter=productive effect. So does that mean sell a lot of non-steel cables? Maybe not. Maybe it's not productive expense. I think the keys are the likliehood of bolt failure in tension or crushing (or decay) failure of wood fibers between the nuts and washers. Do we have enough research on this? Maybe a rigid brace alone reliably achieves the objective.

Reply to post by Scott, on September 12, 2000 at 11:23:14:

I agree, Scott. The cable may not always be necessary, depending on the configuration of the tree system.

This brings up another point to consider, though. Mattheck has promoted a uniform stress axiom, where the tree responds with added wood growth where and when nedded to maintain structural integrity. Smiley has suggested the best placemment for a crotch bolt is above the crotch. From a physics point of view, this makes sense. From a physiological view, is it the same? Has Claus commented on this? I suppose if the tree does not fail too soon, it will then adjust to the new structure and physics and respond with new growth to accomodate the bolt placement.