 Okay, some of the other methods of direct reading of fastener tension, load cells, and of course this PLI preload indicating washer that I described in the washer section, is a mechanical load cell assembly, and those can be used, but of course in a lot of assemblies you don't have room enough to put all that stuff in, because the amount of deformation in them has been correlated to specific tension load in the bolt. Now the Skidmore Wilhelm Bolt Tension Measuring Machine shown on the next page has a load cell in it to give a direct bolt tension reading for an applied torque. Now it's a bench top type setup which can be used at a construction site. You just clamp it on to a beam and check some bolts and then you determine the torque that you want to put in that particular batch of bolts and use that for an installation. This company is here in Cleveland, one of the owners of it, attended last fastener conference we had here. See it's a little jobby, it just has the clamps over here that you can clamp it on and it gives a direct reading of the number of pounds that you've put into the bolt, so it is a way of determining torque. Now here's another one that you don't think of it as being a direct reading, but it is, and I use this one, I'll cover it later in the rivet section, this is a high lock rivet or lock bolt, it's actually a lock bolt rather than a rivet because the shank does not expand on it. But it has a, it's a blind type fastener for installation and you have a hex key that holds it in place. While you crank the nut on and when you get the nut to the proper torque limit, it is not shear so that it snaps off, breaks off. So that is a soft limiting type that they have determined the right diameter for them so that it will break off at the torque that you want. Now here's one that I mentioned earlier, the DTI bolt and the guy who has that company was at one of our Bolting Technology Council meetings. This is a color coated type bolt. It has a little gauge pin that's threaded inside it and then it has an optical absorbance cell near the surface and as the cell changes thickness it changes color. So as you elongate the bolt it pulls this gauge away from the cell and gives you a color coated load indication so you torque the thing until it shows red or whatever and you're alright. But of course the problem with a bolt like this, you can imagine how expensive it is compared to a hardware store bolt and it's going to be big enough that you can drill the center of it to put that stuff in so the minimum size for it is a half inch diameter. Now we move on to design criteria and this first line is one of my pet peeves. I think that we don't spend enough time looking at something before we design it. We should think about it, look at it first. Now of course working with the research people around here, you usually do designs by an iterative process because when they come to you about something usually they don't know and I'm not saying this as a degrading remark. They don't know exactly what they want so they tell you what they think they want and then you take it from there and usually what happens is you by iterative process you come up with the actual requirements and that's something that you should do on any design. You should sit down first and look at it, decide what you really need and then look at the accepted design practices from both the layout and analytical standpoints to see what you should do. Now here's one of the questions that comes up sometimes is diameter versus length on fasteners and we're always faced with decisions on do you use off the shelf stuff or do you custom design it? Well, a good way to look at it is to check and see what's available first, see if you can build your design around that without having to buy special components. And so one of the things that you look at is the length to diameter ratio of fasteners because if you want to use a 10 inch fastener that's a quarter inch in diameter and you don't want to make it out of threaded rod you're in trouble because you're going to have to get one as custom made. Usually the L over D ratio is up to about 12 and it's limited somewhat by the capacity of the automatic screw forming machines because you can't put a long skinny fastener through there and do it on an automated basis. So we have a table here that lists common fasteners availability. Now these are industrial fasteners not aerospace and the one asterisk represents the stock sizes of maximum demand. So you see if you're in this area, if you need a three-eighths diameter with an inch or inch and a quarter length you got it. Two asterisks represents the ones less frequently used so if you want say a quarter inch and an inch and three-quarter length you might have a little trouble. All the rest of them are considered specials so if you want a quarter inch by six inch look how deep of trouble you're in down here because you can't get it unless you pay special for it. And I know we had some fasteners on a job that we did here one time that were quarter inch stainless steel that had to be I believe five and a half inches or six inches long. We had to pay special for them and took a long time to get them. Now when I say these are industrial type fasteners on the aerospace fasteners the lengths there are graduated in sixteenths. The different you can specify a dash number that gives you the grip length of the fastener in sixteenths but just because somebody shows it in their catalog doesn't mean that they have it either. So if you need something that's an oddball type length to diameter you still are going to have to pay special for it. Now here's a little handy dandy thing that once again I got from one of these guys at Martin when I worked there and I've never seen this anywhere either is a way of calculating the number of fastener diameter. I put it in my fastener manual so a lot of you guys have already seen it but in order to calculate it this is for inches of course you take sixty thousandths plus thirteen thousandths times n where n is the number of the fastener and that will give you the total od of it. And the example I give here is a number eight fastener you take sixty thousandths plus thirteen thousandths times eight gives you one sixty four. So if somebody says I got a number six or a number four you can calculate the decimal diameter of it directly that way just by keeping that in mind of course the number ten is easy because it's one thirty plus sixties and one ninety. And for those of you who haven't seen them before they even have number twelve fasteners which I believe works out to be point two one six or something like that. The automotive industry uses them some and I ran into some of them one time and I couldn't figure out what they were way back years ago until I figured out it was a number twelve but they're not a normal one. Now clearance holes for fasteners for shear applications the clearance should be minimized and ideally the holes should be matched drilled and the material thickness and fastener strength should be sized to make the fasteners critical and bearing rather than shear. That means that if you pull and shear on the joint that the fastener is stronger than the material it's in so it will elongate the hole so it can load up the other fasteners. And in tension applications you don't have to worry about that if you can assure yourself that you have enough tension in it enough friction that the joint won't move then you can have a looser fit on it. Then your main concern is to prevent the fastener head or the nut from pulling through the hole or something like that or embedding in it. Now on the next page Fred Yarris's group was kind enough to draw me up this little thing because I couldn't find one. Usually you try to steal stuff from someplace else for these to save yourself work but we couldn't get by with it so we had to make one. And this is a little drawing of a joint to illustrate the clearance hole gaps on fasteners and where it gets you in trouble. Now this happens a lot. I'll go to this one over here to maybe it's a little clearer. This happens a lot where you have two pieces that one place makes one piece and somebody else makes the other one. Then you bring them back and you try to put them together. And this is what you get. Now these are the different gaps. See here we have no gap on this. But look what we got up here and look what we got here. So if you pull on that the only way that these other fasteners can load up like for instance here since that one is up against the wall right now. In order for this one to load up the hole has to elongate on here for that one to load up. So this is why that in a real shear applications critical design you should match drill. And this is what the aerospace companies do. They'll take the pieces they'll have a pilot hole in one which is a smaller diameter hole than the hole that needs to be in it at the end. They will clamp them together then they will use that pilot hole to go in with the proper sized drill and drill the hole all the way through both pieces so that it matches perfectly. They're all with the same drill. And then you put it together and you don't have this problem. Now here's another one that we can run into trouble on is mixing of the thread and material types. And this happens in design sometimes because you can have say 300 series stainless steel fasteners and you can have 8286 stainless steel fasteners and you look at them they look alike. The one has a strength of usually of 160 and the other one has a strength of 70. So you can get in trouble with it. So if the different sizes have fine or coarse threads on the same diameters or I mean if you have them with the same diameters with the fine or coarse threads or metric threads then you're in real trouble because to a mechanic all these fasteners look alike and this happened on the CM1 job if I recall we had one that the guy couldn't figure out why it wouldn't go in the hole. And it was a metric course when I had him get a gauge and gauge it and we had inch stuff around there too so and the rest of the metric stuff was fine thread I think and this one happened to be a course so this is asking for trouble because if something won't fit somebody's going to try to make it fit and they put it together. Now we covered the different strength levels and the fact that 300 series and 8286 look alike. Stainless steels look alike and even different platings on materials can be dyed to where they look alike. So next we go to the selection and positioning of the washer and you got to pick washers that are large enough to distribute the load under the head or the nut without exceeding the compressive yield strength of the joint material. So you want a hard washer and a smooth one so that you know what your coefficient of friction is going to be and if the internal diameter of the washer is much larger than the fastener then you better try to center it to make sure that they will fit. Don't do one of those deals like I've seen people do before where they stack up a whole bunch of washers and then they got to jiggle them around to get them to fit under the head and you might wind up with the thing embedding in the material on one side and on the other side it's hardly loaded. Now shear loads on a fastener group. This is something that I gave you a lot of verbiage on this to help you go through the stuff on your own. And so I'll just kind of hit the highlights on this. Number one on where you have a pattern of fasteners, the first thing you want to do is determine the centroid of the pattern by picking X and Y axes and using unit areas times its distance to get the centroid. And although it's not a good idea to have fasteners of a different diameter you can use them in this type of analysis by ratioing the diameters. For instance the one I gave here, if I had eight bolts of a 12 volt pattern that were three-eighths and the other four are five-sixteenths, you can ratio the shank diameters and you use one for the one that you have the most of and use the stress ratios then to give you a factor for the other one to use. This way you can calculate the pattern CG and get the loads on it. Okay now, in a lot of cases you'll have a symmetrical pattern so you're okay. But after you find the centroid then you can get these sigma R squares for the fasteners which will give you an equivalent moment of inertia if you will like calculating bending stresses. So we can move over to the figure and I think I can talk you through that better. Here is a bracket that has an eccentric load on it, here R. Okay now to get, and it's loaded just in shear, we're not putting any tension on it. So we have to transfer that to the CG, of course remember in strength of materials you transfer a load to the CG, you have a direct load and a moment is what you replace it with. So you have this as your direct load is just taking R and divide it for the number of fasteners. It gives you a load there. Now you get a moment R times this value E which you have to react. Now the way that you react that you take these R values which is the distance, this is the centroid since it's a symmetrical pattern. So you have four R values measuring here, here, here and here that are the same and then you have four more that are the same from here to here, from here to here, up to there and down to here. So now you take those and add them up. So you have four times R1 squared plus four times R2 squared and that gives you your equivalent moment of inertia if you will. Then you can find a load on the fastener by taking the moment times the radius to the one that is farthest away and that then over the sigma R squared value that you calculated using those values and you come up with another load. Now you take those two since they're both in the shear plane. You combine them vectorially to get a result load P for a total shear load on the fastener. Then of course that takes care of the shear loads. Now if you look at this value, this also would correspond to like a torsional formula the TR over J in which the sigma R squared is the R sub n squared is the equivalent of a polar moment of inertia J except that the load that you get here is in pounds and so then later on if we have tension on something like this we can combine it and get the total load using stress ratios. Now on edge distance and fastener spacing this is something that's violated a lot. In fact we put out designs around here before that I have been very disappointed with because somebody used practically no edge distance on stuff. We won't mention any names but Ron knows a guy that did this a few times on me. But here is the edge distance and fastener spacing and these are nominal ones. So this is kind of what you shoot for. 2D nominal where D is the diameter of the fastener. 4D spacing between fasteners. And the aircraft companies usually use a 2D plus 30 thousandths on their stuff just to give you just a little more edge distance in case you run into a problem. Now one of the things the questions might be asked well if you have a shear lug it doesn't have 2D. No they're custom designs because they're usually pretty thick and you go in and calculate hoop tension and shear tear out and that types of things on a lug. And that one is covered, I covered it in the chapter I wrote for that textbook. It's not out yet because that was on fasteners and shear. And Shigley and a few other people also have coverage on shear and lug design. When you're talking about lug you're talking about a crank that is fairly a crank type thing that is fairly thick. So and usually you have since it's a rotating type joint it does not have a large edge distance but it's got thick walls. Now here's something that I use to illustrate one of the other fallacies that we deal with in the engineering world. The development of bearing stress allowables. Bearing stress is the normal way of doing it. You take this sheet as thickness T. Now this represents a semi-circular, there would be a fastener fitting in that hole and this represents the way we're coming up with the bearing stress. Here's what you're actually doing. Because if you have a fastener in this hole pushing the maximum stress is right here. So this represents that maximum stress here. It's zero here because you're not putting any stress on it there. So what we normally do and see mill standard 13-12 which will be covered later on in here gives all the different methods of testing of fasteners. Well what they do they put the fastener in the material and they test it to failure. When they get it to failure whatever it failed at for a given diameter they divide it by the diameter times the thickness of the material that's your normal bearing area and say that's the bearing stress. So if you look in Mill Handbook 5 or any of these books on bearing stress allowables you will see that they are way above tensile element and tensile yield because they're a fictitious thing. What they are they're a value that has been verified that you can use it for calculations and get by with it but it's actually not a true stress. So if you don't have bearing stress allowables for material since you see that these are proportional P1 to the compressive yield equals P2 to the ultimate and so on you can come up with these just by taking one and a half times the compressive yield or compressive ultimate of the material. Now that is a conservative figure and because the actual test value will run around 1.7 for most of these materials but Mill Handbook 5 if they didn't test to get the bearing allowables in a lot of cases they'll just take one and a half times the tensile element or tensile yield and slap that in there for the bearing allowable because they know it's safe. All right grip length and shear head and tension head on the passenger. Now grip length is a very critical thing for shear design because that is the length of the unthreaded portion of the fastener and when you have it in shear and you try to keep have no threads in the hole. So this is the thing that you do here and you're supposed to size the fastener such that this doesn't happen. So you put a washer under the nut to allow tightening without running out of threads. Now the airspace fasteners the MSNAS AN nut type are available with shear nuts or heads or tension heads and nuts to save weight on design because if you're designing in shear you don't need to have that much tension so therefore you can go with a thinner head or a thinner nut. So we have illustrations of those in the next figure. Here is the grip length illustration. It's the bottom of the head to the end of the threads and here is a shear head for a same size fastener. It's an eighth of an inch thick. Down here it's 530 seconds for a tension type and notice two specks here that are called out which those of you that are familiar with the fasteners this is for J thread shear of the mill S8879 and this is for the class 2 or 3 in the standard threads. Here is a shear nut and a tension nut. Well you see the shear nut is pretty thin 203 versus 284 for the tension nut. So if you have a joint that is primarily shear you can put in a little nut like that and if you're using several hundred of them it saves you quite a bit of weight on a airframe. Now here's another thing I keep coming back to. Avoid tapped holes. We covered the tapped holes and the type of taps and here's some more reasons for avoiding tapped holes. Cost. Drilling and tapping a hole is expensive compared to drilling a clearance hole for a nut and bolt assembly. Inspection. About the only thing you do with a tapped hole is a go-no-go gauge and a minimum thread diameter check just by running a pin through it and the root radius you can't measure very well and since there's no such thing as a UNJ tap the root radius is not rounded. If the hole is blind it will have burrs, shavings and everything else and you're just stuck with it. Now here's the other type of design that you need to look at is tension loads on a fastener group. And at the time that I did this one I couldn't find one anywhere in anybody's book so I had to draw this one up myself but it didn't get too fatigued during our scanning so I guess it's alright. And here we have eight fasteners on a bracket that has two different loads under it. It has a direct tension load P1 and it has a shear load P2 which also gives you a bending moment. So what you're trying to do is get the total load on all of these fasteners using the different loads that we have there. Alright? The moment from the load if you, now here's another thing that's different from the other one is where do you measure R from? R is measured from the healing point for your sigma R squared because if this thing goes into compression over here then you're not getting anything out of it for your tension load so you can't use those two fasteners to carry the tension because they're in compression they're not going to help you any. So what I did in this case since it's a bracket and this is a plan sticking out I said okay this thing is hard up to here so I'll measure my Rs from that point to the right so I only have for my sigma R squared I only have six fasteners in it but then for the total shear I'm using all eight of them and for the total tension I'm using all eight of them. So in doing that you can calculate the sigma R squared you divide the load I better leave this up here from my standpoint here for a moment you can divide the load by eight to get the one shear loads then you can calculate the moment which is the P2 times H then take the R7 which was the one further stout and use the sigma R sub n squared and you can get a load then a tension load from the moment you have a P sub t on there which you divide by eight to give you the additional tension load and you can go in then and calculate the total load in tension then you have the shear load which was the P2 over eight and you can take those two loads now and go in and use stress ratios and calculate the margin of safety on the fasteners for the total loading see here was a better print showing the one for the P sub m value where you actually get in the moment was the P2 times H and that times R7 over the sigma R sub n squared now the tensile load the preload that you're putting on these fasteners has to exceed P or you're in trouble because you don't want any joint loosening combined shear and tension loading now on this you have you get all your summation of loads in the shear direction you get all your summation of loads in the tension direction and then you could use a motor circle and work with it and get the principal stresses and that type of thing and calculate out an allowable and a margin of safety that way but it's easier to use these stress ratios because that's doing the same thing so what you do is you get two factors you get a R sub s or R sub t here which is the actual shear load over the allowable shear load for that fastener now in this case you can work in pounds you can work in stress either one you want to as long as you're consistent in your units so you get a factor there you get one from tension the actual tension load over the allowable and then you get a margin of safety which takes the actual load over the one you calculated here minus one to give a margin of safety now what happens with these these values when you combine them you get two values that had better be less than one for each one of them because you don't want either one of them to be greater than one or you're in trouble on the design so you have these and they have exponents x and y now it depends on your degree of conservatism as to how big an exponent you use for those because of course the bigger the exponent goes the more unconcervative you become because the sum of those two have to be less than one in order to have a positive margin because margin of safety as a safety factor of one if you will so a margin of safety of zero is a safety factor of one so therefore if people say well gee I got a margin of safety of .03 on that part well that's still good because that's 1.03 safety factor wise and that's the way the airspace industry has been doing it ever since Glenn L. Martin so here are these curves that you can use and it depends on how conservative or unconcervative you want to be now for if you're the belt and suspenders type and want to make sure everything's all right you can use the straight line version here which is just uses no exponents at all and calculate the margin and that one is a lot safer if you want to get more unsafe you can go further out on these by squaring and cubing these ratios and that will give you a better margin of safety for a given load now here's one that is another one that has always bothered me because in school I never did like the way these professors went through horizontal shear stress and said in the determination of this is an exercise left up to the student so anyway I went through and developed this for the lecture that I give on fasteners and shear because it had always bothered me that nobody had explained it very well and of course when you're looking for explanations and strength materials you go back to the basics back to the real source, Timoshenko so I found this in an old Timoshenko book in which he explained it and the book was old enough that he wasn't working with bolts he was working with nails and tuba floors but nevertheless the principle was the same because when you have two pieces that you want to fasten together so that they act as a beam you have to have enough fasteners to carry the horizontal shear stress for that to happen so this is a method of calculating it and this is the VQ over IB shear stress and so I set up a little problem and worked through it here and these are the dimensions which I think most of them are given on the with the figure I believe yeah now this is a the type of beam and I just came up with a kind of an artificial type thing to illustrate the point you have a 400 pounds per inch loading and it's 50 inches long and it's made up of two one inch plates and you're wanting to hold them together with bolts and you're going to have two at a at each spot so you want to know how far apart your rows of bolts need to be how far can you go and still hold the thing together so that's this E is the spacing here because you see what you actually get when you apply the moment then you have the horizontal shear surface here which in this case is the neutral axis of the beam and you need to calculate that stress and determine the bolts alright you get the reactions to the beam and then you get the moment it's a uniformly loaded beam so it's WL squared over 8 and then you determine a value here and you also have to check bending stress to make sure that your bending stresses are right even if you do carry the shear you still have to hold it in bending so I just took a guess at the diameter bolt and said well I'll use a half inch bolt in this and see how it works out and then I'll calculate it so if you go into this the VQ over IB remember the V is the vertical shear at the point Q is what's called the statical moment which is the area above which you're wanting to check the stress above that shear plane times the distance to its centroid that's a statical moment then so the Q here was the I calculated was 3 inches cubed because it is a area times the distance which makes it cubed I calculated the moment of inertia in this case I left out the diameter of the holes on this because I was just doing it rough and of course one of the things that you should do in the final calculations you actually deduct for the diameter of the holes in order to get the proper moment of inertia then I went in and said for no bolt hole reduction I'll have this stress now that will be across that shaded area back in the figure there which was 6 wide so it was 6e so I solved for the number of pounds that I would have that I'd have to react at that point alright if I take 2 half inch diameter grade 5 bolts good for about 10,500 pounds a piece and I'd divide this total load into that and solve for e I get 2.8 inches maximum spacing between row of bolts so then I went back and said okay I'll use 9 sixteenths bolts with a clearance hole and now I'll deduct for the holes that I'm taking out and calculate a new I and then go in and calculate the shearing stress take that to get me a value involving e and then solve for e using the higher allowables for the 9 sixteenths bolts 8.94 inches for row spacing now you could optimize on that and do all sorts of things but what I was interested in here was just showing how you would do it because none of the books that I had actually strength material books showed that the way it was supposed to be I didn't think so now you still have to go in and check for beam bending and varying stress calculations and notice also that thin structures would have to be checked for inner rivet buckling because if you have thin sheet and your fasteners are spaced too far apart the sheet can buckle in between fasteners under compressive load and you've heard that statement about a little knowledge is a dangerous thing I was telling a guy that this could happen one time at Martin and he was taking strength materials so he said nah that can't happen so I went to see one of the old timers to find out how to get out of it and he said we'll just go up and tear a page out of his notebook that'll prove the point because you see if you take a page and you pull it this way the sheet will buckle between holes before it tears up so I did that never had any more trouble he went back to his strength materials book so that's how to carry the horizontal shear loads now we go into bolded flanges with O-rings now granted that is a science within itself but we'll cover it here just to let you know that you have to do this with bolded joints O-ring compression in a flange is usually just a small portion of the total bolt load and of course the O-ring groove is sized to give a specific range of compression on it when you go metal to metal on the flanges now for most O-rings this compression value is like a minimum of 10 to a maximum of about 30% of the unloaded cross section diameter and of course the flange surfaces have to be smooth to assure sealing without tearing up the O-ring and the fastener spacing must be close enough to keep the flanges from separating that's one of the things you have to watch about now granted in most cases it's not a problem and of course the this has just a general design practice you machine the O-ring groove in the cheaper of the two mating flanges because if the machine cuts the groove too deep the parts scrap so you want to make sure that it's in the cheaper of the two flanges so you can throw it away if you need to and if you need a dovetail groove to hold O-ring in place during assembly or disassembly that also can be machined in now here is a generic O-ring joint and the there's the O-ring normally the the only thing you got to worry about is having a smooth enough finish on this mating surface in that area that it doesn't chew up the O-ring and have enough fasteners to keep the flanges metal to metal now if you go to bolted flanges with flat gaskets then you got another problem you need to squeeze the gasket to seal it but on the other hand you don't want to squeeze it so much that you yield it in compression and run it so now you have to look harder at the amount of load that you're putting in with your bolts now a lot of gasket manufacturers would give you a pounds per linear inch or something for a flat gasket so that you know then by your bolt spacing how much you need to put in the thing to seal at least that gives you a minimum load that you have to have and then of course you have to look at the compressive yield of the gasket to see whether you're putting too much load in or not usually the best thing to do on that is get the information from the manufacturers because they know their product well enough to give you the proper values that you can use and of course you have some things in subsequent sections on what to do where you have gaps on flanges now here's a regular flat gasket joint and one of the things you normally do with gaskets too if you're in the automotive world you use some sort of a gasket cement sealer or something of that nature on them to stick them in place while you're putting the joint together and we have loading curves for the flat gaskets in the appendix which you get one of these days here in the near future and the flat gasket joint design Bickford has quite a bit more coverage on it and so between that and the manufacturers chances are you can come up with enough information for that now gasket loads in flange joints leaks usually start at the point of maximum flange bending which is midway between adjacent bolts where the gasket is not compressed enough to seal a lot of you have run into that in the past with valve covers on cars they don't have enough fasteners in them and you have caskets so you take them down and the thing will bow and leak in the middle and so you have you put cardboard under your car and so to increase the load at the midway point you can look at three different ways of doing it one is to increase the number of bolts increase the flange thickness and increase the initial bolt torque so those are three things that you can look at all of which have their advantages and disadvantages the increase number of bolts since deflection is proportional to the cube of the span between bolt centers that cuts way down on the deflection of the flange and so so adding a bolt at mid-stand gives you a cut of eight on the deflection and increases the gasket load the only thing is you increase the cost because you've now added another bolt another bolt hole increasing the flange thickness now since flange deflection is inversely proportional to the cube of the flange thickness you double the thickness it decreases the flange deflection by a factor of eight so that is a good thing except that you increase the weight and the cost of materials so that's another thing that you have to weigh now increasing the bolt torque is the cheapest way of doing it but if you increase it to a certain point the flange can bend in the middle because you're compressing it down under the bolt and it will allow it to bow up in the middle of where it will leak worse so particularly if you have a soft gasket like the Kirk gaskets you get leakage so and if the bolt is near the yield point a further increase in torque can't be made unless you use bolts with higher strength so one of the cheapo ways that you can do on this is put extra diameter type washers under the bolts to spread the load out just a little bit sometimes that will stop them from leaking but that's not something you'd want to put in an original design now getting into bolted flanges for glass windows the reason I put this in is this is a special one and we've used it around here on designing of windows for pressure vessels because normally you don't think of a window needing much in the way of gaskets and they more or less just slap them in and they're done with it they're in cameras and things of this nature but where you need to use site gauges or in our case actually in CM1 you use cameras through glass now you end up with an optical quality window that costs several thousand dollars that you need to make sure that nothing happens to it so the way to make sure nothing happens to it is that you kind of pad it all the way around with rubber to keep it from touching the metal and then of course the one of the things about it in ours in particular was that the faster design becomes a balancing act to seal it without overloading it because you don't want to overload it and the other thing too is glass is so brittle the thing that causes it to fail of course is surface imperfections and you can't afford to scratch it with anything so it has a coefficient of thermal expansion about one sixth that of metal so if you put it in and you have a temperature change now you have to put enough padding around it of some sort to allow it to be compressed by the material around it or expanded and so on without leaking so that's a balancing act to push it in flat rubber gaskets with a bumper strip around the outside of the window to keep it from direct contact with the metal and then the balancing act is to seal it so that it won't leak but yet not compress it too much so we have a typical design showing in figure 35 and now this is a model of what we actually used in CM1 except that the one thing that I didn't show just for clarity was the O-ring that we had there but you see I'll use this one it's a little clear here is the window you have a bumper strip around the outside now this is something the sealer is just to keep it when you drop it in the socket that it's in or the well there to keep from touching it you have rubber gaskets on the bottom you have a round gasket there and then you have one on top now what you're doing here you're going metal to metal with this top flange now you have to use the tolerances of both on machining of this and machining of this surface in order to make sure that you can put that window in there size your gaskets properly and in some cases you have to grind them to get them to the right diameter I mean right thickness because the rubber is not close enough tolerance put it in there bolt it all down and seal it without hurting anything so that is a special design within itself and of course with your bolts you have to make sure that they're strong enough to go metal to metal and load the thing up without over stressing the bolts now the effect of friction in a clamp joint in most cases friction forces between clamp surfaces are not included in the shear calculations the reason being that's too hard to determine what they are because if you've got oil or grease on the surfaces your friction coefficient could be real low if you've got striations of some kind on it it could get real high but you really don't know what it is so for that reason you normally don't include the axial force of the bolt times the coefficient of friction as a shear capability that you have now there are a few cases and the next page will show there the actual friction forces that you can get you see you have a bolt preload of P and so you take that times the coefficient of friction on this surface and this surface and you've got two forces here that two of these N forces and the N is the friction load which is normal load times the coefficient of friction by definition so in some cases in the construction world they actually use this friction load when they are doing the joint calculations because they count on it but it's not something that you would normally count on because it's too unpredictable now the compression cone of a bolted joint we covered this earlier there in the stiffness section but in the appendices we do give more stuff on it and the here's something that I alluded to earlier the bulk joint relative stiffness calculations and most of the ordinary designs are not a big requirement it's just that where you have say small areas that you would want to like for instance if you had some now one of the things that would be a real problem if you have a bushing type thing around the bolt or a spacer or something like that then you better go in and check out in real fast because you could get into trouble but if you have say steel and you're using steel bolts chances are the joint is going to be stiff enough that you don't have a problem with it you can look at it and take the method of least work on it calculate a circular model for the stiffness if that is satisfactory then go no further now if you were boling say through all soft aluminum, copper something like that then you would probably want to do stiffness calculations to make sure that you're not in trouble but in most cases you can get by with a minimal amount of joint stiffness calculations now boling of dissimilar materials as I mentioned earlier in the Centaur case where you went from room temperature to minus 300 or something like that with dissimilar materials there you got a real problem because differential thermal expansion and contraction of the materials because aluminum is something like three times I believe isn't it the on thermal something like three times the coefficient of steel is and copper is way up there so if you're boling up a copper joint and you have a temperature change you got to be real careful on it but then the other thing that you needed to watch for is the galvanic corrosion because unless the mating surfaces are insulated from each other and that was one of the reasons why the magnesium has kind of gone out of vogue because how do you insulate it satisfactorily that over a period of 20 years if it's using the airplane component that it's going to stay insulated because a lot of these organic type things that they use for insulation the sealers that they put around rivets bolts and stuff like that on airplanes over a period of years can deteriorate and moisture and the other problem with the fasteners on an airplane is that most of them you're looking at heads sticking out so if you have a crack and it's starting at the edge of the hole it has to come out quite a ways before you can see it so that has caused a lot of problems there. The other thing that you need to look at is the yielding of softer materials because if you are say using a high strength bolt in aluminum and you crank that up too much you can actually yield the aluminum in compression under the head without doing anything to the bolt so and then of course you got to check the strength at the temperature extremes because like for instance aluminum falls off drastically at only 250 degrees for a steel most steels will go up to 700 and up before they start falling off in strength so if you were tightening up an aluminum joint and you ran it up to say 250-300 degrees you could get yielding real easy under the heads of the bolts. Now maximizing the effective length of fasteners. Of course when we discuss the stiffness ratios the effective length of the fasteners was mentioned and this is important on the differential expansion contraction so it may be necessary to add a spring or a Belleville washer under a bolt head to increase its effective length enough by the design so that it won't loosen up and the deflection of course is the PL over E so you increase L you're doing alright on it. In fact on the exhaust system on some of the Ford trucks they actually have a big spring on the bolt that holds the flange to the catalytic converter as put on there I think to take the temperature of the differential that you get between the materials on because you can go from room temperature up to about 1300 degrees or something like that on them. Okay we'll take a break here for now and come back up with the match drilling of fastener holes which is an important one.