Grade 8 shear bolt confirmed!

Junkman · Dec 22, 2003

( The grade is the number of hash marks + 2.

grade 2 = no hash marks.
grade 4 = two marks.
grade 5 = three marks.
grade 8 = five marks.


Junkman were you asking me personally or was this posters information correct? /forums/images/graemlins/grin.gif )

No, but you did point out one fly in the ointment that I had originally missed ..... if the grade is the number of hash marks + 2 then a grade 8 having 5 hash marks + 2 would equal a "grade 7".... darn.... this new modern math is hard to understand... /forums/images/graemlins/blush.gif ......

Villengineer · Dec 22, 2003

DO NOT TRY THIS UNLESS YOU WANT TO FIND THE WEAKEST POINT THAT ISN'T THE SHEAR PIN. If you do not get the notch in the shear area of the mechanism you are still trying to shear the full diameter of the Gr8 bolt. If you want to use trial and error method then start with a Gr2 and if it fails under normal use then step up a grade, don't work the other way. Plus why spend more money initially on Gr8 bolts and then weaken them, which requires your time and my time certainly isn't free. It would be vary rare for a manufacturer to use a Gr8 as a shear bolt if for no other reason than they cost the manufacturer more $ than the lower grades. The only reason that I can think of for using a Gr8 as a shear pin would be if there simply wasn't room to use the appropriate diameter Gr2.

keeney · Dec 22, 2003

One thing I learned from doing a lot of technical writing - always leave in a minor error in your documents just to check and see if anybody is actually reading them /forums/images/graemlins/wink.gif

Grade 8 = 6 marks, not 5 as has been pointed out.

I will add from my experience working in the Toro test department that the big companies spend a lot of money DESTRUCTIVELY testing the various designs to gather information about what happens when a cutter hits something hard like a rock.

Yes, it is really fun to run a 1-inch diameter steel "stake" up into a mower deck from a reinforced spring-loaded hole in a concrete floor and see what shear pins break or not (and what else flys off and how high and far it flys) when the company is paying for it, has a concrete bunker to perform the test in, and the information is required by federal safety regulations.

Doing these experiments on your own just for the fun of it can be a very expensive hobby.

A lot of times, the overriding criteria on a design might be safety, not just cost or convenience. In the above example, the safety requirement of nothing flying off to injure the operator might mean that the machine is intentionally designed to fail in some other way to absorb the energy. On the smaller walk-behind mowers, for example, it is common for the crank shaft to bend as well as the shear key to break. Due to the assymetric geometry involved, there is no way for a shear key alone to absorb all the energy of a full speed, dead-on blade strike against an immovable object.

If the crank shaft was made stronger, then the blade mount might fail instead. Or the blade itself would shatter or something else bad.

(The 1" steel bar ends up getting bent at about 10 degrees and has a 1/4" to 3/8" deep divot in it).

- Rick

nomad · Dec 24, 2003

(
I will add from my experience working in the Toro test department that the big companies spend a lot of money DESTRUCTIVELY testing the various designs to gather information about what happens when a cutter hits something hard like a rock.
)

Did I understand correct? Big companies like Toro destructing the cutters a lot in their labs? I don't find such tests logical tests (unless it's testing/cracking a component or two like the shear pin only) because the crack probability in a cutter with hundreds of components is very big if the whole cutter is under crack/destruction tests. I think they are probably destructing only one of its components or two which are difficult to control their parameters. And, this (testing a component or two) shouldn't cost much money. How many units of cutters did you see they were destroying to fix a design?

Egon · Dec 24, 2003

Back when I was younger and still on the farm we had a shear pin problem on the double tree but Dad just fixed it with a bigger bolt.

Egon

keeney · Dec 24, 2003

They did lots of individual component or sub-assembly tests, but every new model also had to go through some mandatory safety tests that have the whole machine running.

Some examples that I recall:

- The stake test I described where the blade is run at full speed into a pop-up steel bar. Nothing is allowed to fly off beyond some distance/angle.

- Tests involving some number of steel ball bearings (3/8" dia?) introduced into the mower deck. A ring of cardboard is placed around the mower to record where the ball bearings strike. Only some small percentage are allowed to strike the cardboard hard enough to leave a dent more than 1' up at 10' or so radius (I don't recall the exact dimensions).

- For rear-bagging units, one difficult test to pass was the nail test. Some number of pounds of roofing-type nails (of some ansi-specified dimension & material) are introduced into the mower. None can penetrate the rear bag. Sometimes some of the nails would get their heads clipped off and would make their way through some of the bag designs that failed.

- On ride-on machines, they would intentionally tip a machine on a tilt table with a 200lb sandbag belted in the operator's seatuntil it rolled over to verify its actual tip angle. They would do this in all four directions.

- I am unclear on the details, but I also recall some tests where the deck would have to be driven into some kind of fixed obect at the full operating speed of the machine.

Some of the non-safety testing was destructive in that it involves running the machine until it was worn out. A typical non-commercial walk-behind mower, for example is expected to last 1000 hours. Commercial units get 2500 hours or something like that. They had bays and bays of test cells where the machines would run 24 hours a day with some kind of rigged-up varying load simulating actual use conditions.

[ Many of these test cells all had fuel on-tap plumbed from a tank on the roof. Some interesting safety issues with that. ]

A lot of the durability testing involved cutting actual grass. They would hire crews of labor to cut grass. Often they would mow a large field at one setting, then lower the deck and cut it again. It was tough to find enough grass to cut because they didn't want to take business away from the local commercial mowing services who were their customers.

- Rick

Villengineer · Dec 24, 2003

For brand new product lines it's not uncommon for the company I work for to make 10 or more prototypes. We usually do a structural destructive test on at least 2, at least one will be cycled first to the design parameters and then to failure if possible, another one is usually sent to our various design facilities to be reviewed, and the rest usually are given to select customers to operate for anywhere from 6 months to a year. The units that are put in the field are checked regularly and when returned completely disassembled and examined. I know that many companies, over many segments of industry, do similar testing.

nomad · Dec 25, 2003

Such tests being done to the prototypes before fixing final design are being done because of traditions in engineering/manufacturing business. These tests actually aren't very necessary. Why? Because, today, all forces (hit by nails, rocks, etc) can be calculated (approximately) on "table" before making a product (by using a force analyse software coupled with a soild work modelling.) For example, a rock hitting a rotary cutter deck can easily be modelled and the stress flows at every point of all cutter components can be simulated on the computer screen. With this computerized technology available (unlike old days), digitization of whole cutter and design of prototypes are easily possible today and that "prototype on the screen" will be almost exactly same of "real prototype out of the screen". So, why still testing the product by using nails, rocks, etc if these aren't necessary anymore? In my opinion, this is because of old engineers in the field who are used to old traditions. If there are defects in castings, human errors in installing the products, etc., then each product coming out of the factory has to be tested (by using rocks, nails, etc etc) and these tests are actually quality control tests (not designing tests before fixing the cutter.) But doing such quality control tests to can't be done because it will be very expensive if you test each of tousands of products by nails, rocks, etc etc. If they are searching the value of an emphirical parameter in a cutter component by hitting a nail onto, say, its deck, then this can be done by a theoretical experiment using only one deck and one nail in a laboratory with a complex facility. If their use of hundreds of nails is a statistical test work, then the test results may correlate well only with clean fields like the golf turf fields where the nails are a few in. Anyways.. It's still good to see a new designed/made product in the real field to feel sure the product is working as designed on the computer screen. But, I still don't think they are very necessary. Ok, this is a debatable topic. Villengineer, so, you are in manufacturing field? What product(s)? if I may ask.

RoyJackson · Dec 25, 2003

"Such tests being done to the prototypes before fixing final design are being done because of traditions in engineering/manufacturing business. These tests actually aren't very necessary."

In theory, you're correct. In practice, manufacturers do "real world" testing to ensure things fit, or there's no tolerance stack ups or hypotheses (assumptions) made during the design stages were correct.

CAD/CAM can save money and some testing, but don't think it's the do-all/end all it's being advertised as...

The real world has more variables then can be modeled on a computer. Real world testing is to reduce risk and determine robustness of a design.

nomad · Dec 25, 2003

(
CAD/CAM can save money and some testing, but don't think it's the do-all/end all it's being advertised as...
)
CAD (Comp. Aided Design) is only a transformation of formulations into the computer language. So, CAD/CAM is nothing in a development process of a new product. I'm talking somethings different.

(
The real world has more variables then can be modeled on a computer. Real world testing is to reduce risk and determine robustness of a design. )
We are saying same things that real world has many more variables. Much bigger computers are needed to include ignored variables of real world. These ignored variables in mathematical formulations are so many that you can't make so many real world testings for each of many parameters/variables. You will have to make one test of a mower against rock, another test against a nail, another test against the wetness, etc etc. Tousands of real world tests... and this can not be done unless you spend A LOT. On the other hand, using computers, you can make many many tests which will not cost you much.. Anyways.. I consider such so-called real field tests only as a "show", to show their lab facilities, to show their designs (at tables) are really working, etc.

Ps: By the way, a mower which will mow a golf field can be better approximated by a computer mathematical modelling better than a mower which will be used in the grass field of a farmer. Golf fields are closer to a mathematical grass field formulation as they are cleaner than farmers' grass fields. So, golf field mowers will require much less test than other mowers.

JimMc · Dec 25, 2003

You have to understand American liablity law. Even with all the theoretical testing, American courts want to see the "hard" data if you want to have a prayer of surviving a wrongful death or negligence suit. We been fortunate thus far, but make no mistake about it, physical testing and the data it provides is the only thing to protect you. I never want to be in the position of facing a family and giving the excuse "well, it worked in theory".

Of course, it is expensive and many companies try to get in and make a fast buck by copying someone else's product and throwing it on the market. But if you want to survive in the long run, the data must be from actual physical tests documented on paper and video. Just saying "we made it just like so-and-so, therefore so-and-so's testing should suffice" or "we tested in theory by computer model" won't stand up for a minute. My advice, use the modelling to get a theoretical "best design" then prototype and test, test, test along with loads of documentation. The Farm Equipment Manufacturer's Association developed tests for cutters years ago and the first thing any court is going to ask for is your proof of due diligence.

nomad · Dec 25, 2003

What have these to do with my thoughts above? You invited the law into a technical discussion. We should not forget that the consultants of the lawyers in the technique/engineering are again engineers/technicans. I guess you mean by "hard data" that should be submitted to somewhere like the law offices "the data" required when a patent/trademark/standart/suchs is applied by the maker who is not necessarily a manufacturer. For example, the patent owners are usually individuals instead of the producers/manufacturers. Anyways, all these are a subject of new topic like "Laws & Engineerings" and we can talk about these if a new topic is opened. Here, I only talking from technique/engineering point of view.

Well, no one is saying "well, it worked in theory" here. Of course, the prototypes will be made. But their so-called Real world tests are not applicable to "Really Real" world. Such so-called real world tests are actually to show/say that "we made it in theory and it worked in the real field with the Parameters we considered in the theory". This isn't the Real world, it's only a theoretical real world which is only close to very clean luxury golf fields only.

Ps: concerning "copy" of products; Shall we discuss about that combining in a "Law&Engineering" topic? The title of such a topic can be "Copy&Universality - Laws & Engineerings".

JimMc · Dec 25, 2003

What do they have to do with your thoughts above?

Did you not post:

Such tests being done to the prototypes before fixing final design are being done because of traditions in engineering/manufacturing business. These tests actually aren't very necessary. Why? Because, today, all forces (hit by nails, rocks, etc) can be calculated (approximately) on "table" before making a product (by using a force analyse software coupled with a soild work modelling.) 

My point is that these tests are necessary. Both as a final proof of theory and as a necessity of law. One can theorize the best product ever concieved and fail miserably in bringing it to market. If a product survives to market, it can still fail from a variety of factors, only one of which is an unfortunate encounter with the American legal system.

Perhaps this thread was meant to be less conversation and more of a monologue?

Please excuse my intrusion.

nomad · Dec 25, 2003

(
My point is that these tests are necessary. Both as a final proof of theory and as a necessity of law.
)
I said it (that you quoted my words in blue), but I see here (I quoted your words) that you are missing my point. Final proof of theory can only and only be accomplished (sp?) by using same parameters that you considered in theory also in (so-called) real world tests. I mean your real world test domain should be so that it should be same as theory test/design domain. Hence, we see many errors later in final products when they are in use by many many farmers in really real fields. As for the laws; I say again this is another story which can discussed after we complete technical discussion (we are now talking about the technology from a global/universal point of view. The laws are local rules which have biased considerations on the technology/science - the history has proven that many times.)

(
Perhaps this thread was meant to be less conversation and more of a monologue?
)
If you consider it's so, you can step back and I can continue on my monologues;-) (joke) - If the dialog conservation also requires other side subjects like the laws, it's ok. I can also join that. A new thread or here?

HayDR · Dec 25, 2003

Namad said "Ps: By the way, a mower which will mow a golf field can be better approximated by a computer mathematical modelling better than a mower which will be used in the grass field of a farmer. Golf fields are closer to a mathematical grass field formulation as they are cleaner than farmers' grass fields. So, golf field mowers will require much less test than other mowers." 

With this premise above, Nomad, you could not be further from the truth. A golf course is more demanding with more operator requirements and thousands of other variables. Just to know mathematical equations and apply them into a design does not mean that the design is functional. It is apparent that you do not have first hand operator knowledge, when it come to cutters. If you do not know what output is desired, then how can you design for anything but to what you think will work. That is like having a map but not knowing where your are located and where you are going.

Villengineer · Dec 25, 2003

( We are saying same things that real world has many more variables. Much bigger computers are needed to include ignored variables of real world. These ignored variables in mathematical formulations are so many that you can't make so many real world testings for each of many parameters/variables. )
We don't do actual physical testing because we ignored variables. We do physical testing b/c we aren't arrogant enough to think that we've been able to make 100% correct models. We use the computer models as a tool to help guide the design process, NEVER as proof of concept. Besides, in our industry it's way to difficult and time consuming to attempt to model and correctly constrain an entire assembly close enough to get accurate data from a program such as FEA. Our computer resources haven't been a problem for years. We currently have enough processor power on each engineer's desk to support it. Additionally, as with any computer program the data out is only as good as the data put in. Therefore since you can not eliminate human error we don't rely on the models.
( On the other hand, using computers, you can make many many tests which will not cost you much )
Are you joking? It takes far more time to completely set up correct constraints for a computer model than it does to set up a physical test. Time is $.
( I consider such so-called real field tests only as a "show", to show their lab facilities, to show their designs (at tables) are really working, etc. )
We don't show off our test labs. They certainly aren't on the guided tours. We physically test because it is the best way to check whether the concept, design, and production are feasable. There are also the legal issue, and like it or not, they don affect what we do as engineers. Theoretical engineers can act like it doesn't matter, but practicing engineers can not seperate themselves from it.

Villengineer · Dec 25, 2003

( Final proof of theory can only and only be accomplished (sp?) by using same parameters that you considered in theory also in (so-called) real world tests. I mean your real world test domain should be so that it should be same as theory test/design domain. )
I'm lost, what in the world are you trying to say here? Any and all testing must be done in accordance with the original parameters, so what's your point? If the parameters are wrong, that's a whole other issue.
( Hence, we see many errors later in final products when they are in use by many many farmers in really real fields. )
What do you consider an error? Do you mean the equipment didn't perform as advertised? Do you mean the equipment broke? Do you mean that it worked, but maybe not as well as you hoped? Do you mean that you dropped it off a cliff and you thought it should have been able to survive the fall. I must admit I don't have first hand knowledge about farm equipment, but I would say that in all the units we repair or replace that somewhere around 75-80% were damaged due to operator error, not poor designs or manufacture.

Junkman · Dec 25, 2003

What really is in play here is the difference in culture between Turkey and the U.S. Here we have many lawyers that will hold a company to a higher standard than any other place in the world for defective merchandise and for not properly testing in real world performance. Computer models are just that ... models... in the real world, performance might not reflect the model that the computer said it will. Hence we have civil suits to contend with when a manufacturer has to defend itself against a defect that should have been anticipated or seen before production. The list of recalled merchandise due to defects or perceived defects is compiled weekly and decimated to the news media. With all these events happening on a regular basis, the U.S. manufacturers would not be able to even get product liability insurance if they couldn't prove that through testing was done and continual product development and testing is continually done.

If a product is being imported, many times it can't even get past the Customs inspectors if they have any doubt as to the safety of the product. That is why automobiles manufactured over seas are manufactured to U.S. Standards that are many times different and more stringent than the standards of the country of origin.

nomad · Dec 26, 2003

( Villengineer says in response to my words using computers, you can make many many tests which will not cost you much
Are you joking? It takes far more time to completely set up correct constraints for a computer model than it does to set up a physical test.)
I'm not joking, but are you? Lets exaggerate by taking the airplane as an example. Most of their design work is done on the computers (big computers) by simulating them on the computer screens. You can play with many parameters there on the computers, but you can not make many plane destruction tests which will cost you a lot..

( Villengineer says in response to my words Final proof of theory can only and only be accomplished (sp?) by using same parameters that you considered in theory also in (so-called) real world tests. I'm lost, what in the world are you trying to say here? Any and all testing must be done in accordance with the original parameters, so what's your point? )
What I am saying by "real world test domain should be same of theory/design domain" is that; Lets consider the stress distribution in a mower in operation is S and lets say it is a function of some set of independent and dependent several variables (a, x, y, z, t, c, l, k m), i.e. S=S(a, x, y, z, t, c, l, k, m). S can be a nonliner/complex functions of these variables. As you know when you make a mathematical modelling these parameters are connected to each other following some rules and axioms/assumptions. And, then if you can't solve this final formula analitically(by hand), you digitize it to solve it (approximately) using a computer. Look into this function. How many parameters are taken into consideration when modelling the mower operation mathematically and then computer? It is 9 (nine). You make the design using these parameters and make the prototype. Now, you will be making the (so-called) real world tests? Well, your real world test domain should be such a space that the number of its parameters of this domain too should be 9 (nine) or less. But the outside "really" real world has many more parameters than 9. So, testing a mower designed by 9 parameters in a really real world with more than 9 parameters actually will fail if one of the parameters, say 10th parameter neglected/ignored during math. formulation/comp. design starts to play an important role. This is modelling/design error if you didn't consider that 10th parameter and such errors frequently occurs. Anyways, your mower / design is based on 9 parameters and you can not make a test outside of this 9 parameter space. Therefore, (so-called) real world tests of a mower are being done by companies under isolated environment (isolating the test area from other parameters which were not considered during the modelling/design) and this isolated test area doesn't usually represent the really real field. An exception can be only the golf field which is much smoother (less parameters) comparing to the natural (farmers) grass fields. I mean you will probably get good results if you make the real world tests on smoother surfaces like the golf fields which are artifical grass fields just like theory/design domains with 9 parameters only. Farmer/end user error is that if he/she uses the mower in a 9-parameter-domain and if the mower fails, then this is operating/using error. If he/she uses it in outside of 9-parameter-domain and if the mower fails, this is modelling/design error or this can also be called "a restricted mower" and again it's a user error if that 9-parameter domain is clearly stated in the user manual. By the way, even testing the mower in a domain of 9-parameters considered will require many many real tests (somethings like 9 factorial = 9x8x7...2x1 = 362880 ! number of tests if there is no symmetry forms in that modelling which can reduce this number..) SO, who/which company can make 362880 real world tests? None of them. They make 30 tests and they call that real field tests. lol. ps: the laws is another subject we can talk in a new thread. All I can say (as a reply to Junkman) now is that imported products are, as a tradition, USUALLY in their responsibilities of importers/distributors who have to follow their own local laws. We know that many companies from Turkey, Italy, China, etc are exporting their products to all countries including USA. Do you think these manufacturers outside follow all USA technical/safety standart/etc laws? No. Importers/distributors in USA do that. Shall I give some examples?

Egon · Dec 26, 2003

Sure hope you fellows get a shear bolt picked out soon. Springs coming and I'll want to mow and rotatill. Hate to relie on empirical methods but if required will have to.

Egon