Case No. IP-2018-000199

The pleaded prior art 113.Airbus relied on two items of prior art in support of its argument for lack of novelty: (1)US Patent no. 2005/0169726 A1 (“McClure 1”). (2)The E-Nut. 114.As pleaded, the E-Nut was cited by way of a back-up. It is said to be a specific embodiment of McClure 1 and the Grounds of Invalidity assert that McClure 1 provides a full description of the E-Nut; in case Kwikbolt contends that it does not, Airbus relies on the prior public use of the E-Nut. 115.In support of its argument of lack of inventive step, Airbus pleaded reliance on McClure 1, the E-Nut and (1)PCT Application no. WO 2004/037483 A1 (“McClure 2”) and (2)UK Patent no. 413,403 (“Wylie”). Application at the start of the trial regarding the E-Nut 116.As I have mentioned, there was an application by Kwikbolt at the start of the trial to disallow Mr McClure’s evidence in its entirety because it should only have addressed whether the E-Nut had been made available to the public before the Patent’s filing date and Kwikbolt conceded that point. From this there emerged a larger dispute at the start of the trial which I must now deal with: what should be understood by “the E-Nut” as pleaded? 117.Apparently the E-Nut has been marketed in various versions and Airbus said that it was entitled to rely on several alternatives as pleaded prior art. Mr Ward took me to the evidence of Mr McKay to show that Mr McKay had appreciated that the E-Nut came in different versions and that Kwikbolt must have been equally aware of this and that the various versions all formed part of Airbus’s case. He submitted that Kwikbolt’s attempt to limit the attack on lack of validity and inventive step with regard to the E-Nut to one version should have been made much earlier. He further submitted that Kwikbolt’s application should be refused for four reasons. 118.First, Kwikbolt must have understood that it would be absurd for Airbus to rely only on the version that did not matter. Secondly, McClure 1 shows that the E-Nut as there disclosed may be implemented in different ways. Thirdly, the Grounds of Invalidity refer to Annex 4 to that pleading which shows the ENut comes in different versions. Fourthly, Annex 2 to the same pleading shows the same thing. 119.Mr Ward finally submitted that if I were minded to limit Airbus’s case as Kwikbolt sought, I should give Airbus permission to amend its Grounds of Invalidity so that the trial could continue with Airbus relying on all versions of the E-Nut. 120.I ruled that Airbus could rely only on the version of the E-Nut referred to in paragraph 3(d) of the Amended Grounds of Invalidity and that I would give my reasons in the judgment after trial. Those reasons now follow. 121.It is not unusual for industrial products to be marketed in alternative models or versions, often many. I take the view that when a party pleads such a product in support of an allegation of lack of novelty and/or inventive step, it is essential that the pleading is entirely unambiguous as to which version of the product is relied on. First, the patentee is entitled to know the case it has to meet. Secondly, this court (and indeed the Patents Court) routinely limits a party seeking to challenge the validity of a patent to reliance on a small number of items of prior art. It requires the party to focus its attention on its best case and thereby avoids a waste of time and expense, not least the time of the court. This is plainly in conformity with the overriding objective. It would undermine that routine approach if a party were entitled to plead reliance on product X without further detail and thereby to smuggle into the case any number of alternative versions of product X. 122.These are the relevant paragraphs of Airbus’s Amended Grounds of Invalidity: “3. A specific embodiment of McClure 1 was a commercial product known as the ‘E-Nut’. The Defendant contends that a full description of that commercial product is provided in McClure 1. If and insofar as the Claimant seeks to contend that elements of the operation of the ENut are not disclosed in or directly ascertainable from McClure 1 however, the Defendant will rely on the prior public use of the E-Nut, which was widely publicly available from at least 2003. Indeed, such was the widespread use of the E-Nut that the Defendant contends that it formed part of the common general knowledge of the skilled person as further set out below. The Defendant relies on the following particulars in support of the prior public use of the E-Nut: a.The E-Nut was manufactured by Avibank and was commercially available in at least the United States from at least 2003. The E-Nut is still manufactured and sold commercially by Avibank (see Annex 2 – extract of the E-Nut page from Avibank’s website); b.The trade mark filing for ‘E-Nut’ at the USPTO (Registration number: 3361456), filed by Avibank Mfg., Inc. notes that the first use in commerce of the goods to which the trade mark relates was 11 October 2002. A copy of the relevant page of the entry for the mark at the USPTO is attached as Annex 3. c.To further support that the E-Nut was available prior to the priority date of the Patent, the Defendant relies on the display of the E-Nut at a trade fair in Hamburg in spring 2004 at which E-Nut fasteners were displayed as part of an installation displaying the new Boeing 787 flooring system. Photographs from this exhibition are attached as Annex 4. d.An E-Nut is in the possession of the Defendant and available for inspection on request. The E-Nut in the Defendant’s possession is of the same design as those available before the priority date of the Patent. 4. Further and in any event none of claims 1 to 18 of the Patent is in respect of a patentable invention in that the same was obvious to a person skilled in the art having regard to the state of the art which comprised the following matter at the priority date of the Patent: a)The matters referred to above; b)WO2004/037483A1 (‘McClure 2’); and c)GB 413,403; in conjunction with the common general knowledge. The Defendant refers to the amended claim chart at Annex 1.” 123.I think that any reasonable reader of those paragraphs would take the E-Nut to be a single product in a single version. Paragraph 3(a) would be understood to mean that this product was available from 2003 and that Annex 2 provides confirmation of this. Paragraph 3(c) would be taken to mean that it was displayed at the Hamburg trade fair and that Annex 4 provides confirmation of that prior disclosure. If any doubt were generated by the fact that Annexes 2 and 4 show alternative versions of the E-Nut, paragraph 3(d) would be taken to provide a clear answer: the version relied on is that made available to Kwikbolt for inspection. One version only was made available. In my view that is the single version pleaded and the only version on which Airbus is entitled to rely. 124.Moreover, Mr Vanhegan said that in reliance on the pleading Kwikbolt had not investigated whether there had been prior use of versions of the E-Nut other than that pleaded and had not served evidence addressing the point. Mr Ward accepted that Kwikbolt’s evidence did not cover the question of prior use of other versions. 125.I agree with Mr Ward that Kwikbolt should not have sprung its application to limit the scope of Airbus’s reliance on the E-Nut at the last minute. However, in this instance Airbus either knew or certainly should have known since the service of the Amended Grounds of Invalidity that Kwikbolt was entitled to understand that only one version of the E-Nut had been pleaded. I also think that Kwikbolt might be forgiven for having assumed that it made little difference so far as Airbus was concerned: on Airbus’s case McClure 1 disclosed everything about the E-Nut on which Airbus wished to rely. McClure 1 126.McClure 1 claims an improved blind fastening system. It comprises three components: a sleeve insert, a collet body and a threaded stud. An embodiment is shown in Figures 15 and 16: 127.Mr Ward provided a helpful simplification of Figure 16. The sleeve insert (the putative first member) is shown in red, the collet body (the putative second member) is shown in green and the threaded stud (the putative screw member) is shown in yellow: 128.The collet body has flexible fingers at one end (marked 36 in Figures 15 and 16, with protruding portions 34). Mr Jack explained how the McClure 1 fastener works. The collet body is inserted through the top of the sleeve insert. The collet body is pushed down through the sleeve insert until its flange meets a rim or ledge in the sleeve insert. The rim or ledge is designed to prevent the collet body from falling out of the base of the sleeve insert. The flexible fingers in their resting state are sufficiently converged so that they can extend through the sleeve insert’s base. When the fastener is inserted into the apertures of the workpieces, the converged fingers can project through the aperture of the lower workpiece. 129.The threaded stud is inserted through the top of the sleeve insert and into the collet body, engaging threads on the inner surface of the collet body. Rotation of the stud causes the collet body to be drawn up along the stud towards and into the sleeve insert. The bottom of the stud is forced through the collet body’s flexible fingers, causing them to expand outwards. The collet body is prevented from rotating in the sleeve insert by complementary internal geometry. As the collet body is drawn up, its flexible fingers, now pushed radially outwards, engage the surface of the lower workpiece. The sleeve insert has a flared upper lip which bears upon the upper workpiece. This and the flexible fingers bearing on the lower workpiece cause the two workpieces to be clamped. 130.McClure 1 does not explain how the fastener would be removed. Mr Jack said that it would be clear to the skilled person that this would involve the reversal of the steps set out above. 131.Mr McKay said in his first report that McClure 1 discloses versions of a blind fastener, none of which is removable. As accepted by both sides, for the fastener to work the sleeve insert must be kept from rotating when the threaded stud is rotated. Mr McKay’s point was that McClure 1 achieves this by attaching the sleeve insert to the workpiece. The sleeve does not leave the workpiece on removal of the fastener and therefore the fastener is not removable within the meaning of claim 1 of the Patent. 132.Airbus advanced three arguments in relation to McClure 1 and claim 1 of the Patent. First, Mr Ward submitted that the fastening element of McClure 1 is removable even if the insert is not and therefore the fastener can be categorised as removable. This does not assist Airbus because I have found that “removable’ in the context of claim 1 requires the fastener as an entire unit to be removable. 133.Secondly, Airbus relied on a particular embodiment of McClure 1, one which uses a sleeve insert as illustrated in Figures 15 and 16 and which is shown by itself in Figure 4: 134.There is discussion in paragraph [0061] of the embodiment using the sleeve insert of Figure 4. It expressly contemplates the user “adhering” the sleeve insert into the aperture of a workpiece. The discussion continues in paragraph [0063]. To put the relevant part of that paragraph into context, I will also quote the preceding sentence which refers to Figures 13 and 14: “In FIGS. 13 and 14, sleeve insert 90 is secured to proximal panel 72 such as by adhering base 106 to the bottom surface of proximal panel 72 and the panel apertures are aligned. In FIGS. 15 and 16, sleeve insert 90 is secured in the aperture of panel 72 such as by adhering outer surface 94 to the aperture of [proximal] panel 76. Alternatively, but not shown, outer surface 94 can include a plurality of radially extending elements, such as knurling or vanes, that physically engage with a [proximal] panel.” 135.Airbus relied on the words I have put into italics. Where I have put the word proximal in square brackets, it is because the paragraph says “distal” but the experts were agreed that the skilled person would have understood McClure 1 to have meant “proximal”. 136.Mr Jack discussed that short italicised sentence. He said in his report that use of knurling or vanes would have been well-known to the skilled person as a means of creating friction with the surface of the panel, thus resisting rotation of the sleeve insert. The fit could be loose, close tolerance or interference fit. The choice would depend on how easily one wanted the insertion and removal of the sleeve insert to be. With a loose or close tolerance fit the sleeve insert could be pulled out of the aperture along with the stud and collet body. With an interference fit, a tool might be required. 137.Mr McKay responded in his second report by saying that all of paragraph [0063] of McClure 1 is about securing the sleeve insert to the proximal panel, i.e. workpiece. He also said that knurling or vanes on the surface of the sleeve insert would risk causing damage to the workpiece and should be avoided at all costs in an aircraft assembly. 138.Mr Jack was cross-examined on this: “Q. All of the inserts in the McClure 1 patent work because they are fixed in the panels which are going to be clamped; is that not correct? A. Yes, that would have to be correct. Q. Yes. They are all describing what I think is called by Mr. McClure as an anchoring system, are they not? A. Anchoring system, yes, sir.” 139.Mr Jack went on to say that he thought that the clamping or anchoring disclosed in McClure 1 was very similar to the disclosure of the Patent. I am not sure what Mr Jack meant by similar. 140.In cross-examination Mr McKay did not alter his evidence that the italicised sentence in paragraph [0063] was concerned with securing the sleeve to the workpiece. He was also cross-examined on the basis, as was put to him, that the skilled person would wish to avoid knurling because of potential damage and would think of using a tool and slots instead to prevent rotation of the sleeve insert. Mr McKay said: “Well, yes, he needs to prevent rotation, obviously, for the component to work, as designed. You know, if he was an inventor, you know, yes, he might try to come up with some other version, but I guess that is always an option.” 141.In summary, both experts believed that the skilled person reading McClure 1 would know that rotation between the sleeve insert and the collet body had to be avoided for the fastener to work. The solution disclosed in McClure 1 is to fix the sleeve insert to at least one of the workpieces. There is no suggestion anywhere in McClure 1 that the sleeve insert should be removable. The only option disclosed in McClure 1 that does not expressly require the sleeve insert to be adhered or otherwise fixed to a workpiece is that of using knurling or vanes as discussed in one sentence of paragraph [0063]. In theory it was possible to use a loose enough fit between the sleeve insert and the aperture to make removal of the sleeve insert easy. But as Mr McKay said, the teaching of paragraph [0063] is still all about ensuring that the sleeve is secured to the panel by one means or another. Further, the skilled person would find the option of using vanes or knurling unattractive since it would create a risk of damage to the workpiece. It would also not be removable in the sense that I have construed that word in the context of the Patent. In my view paragraph [0063] does not disclose a removable fastener. 142.Mr Ward’s third argument was that claim 1 is a product claim and therefore sets out physical characteristics. All those characteristics are in the relevant embodiment taken from paragraph [0063] of McClure 1. I disagree. That embodiment is not a removable fastener as I have construed that term, requiring the entire unit to be suitable for removal without damage to the fastener or the workpieces. 143.In my view there are no clear and unmistakable directions in McClure 1 to the invention of claim 1 because there is no disclosure of a fastener which is suitable for removal as a unit. On the contrary, the skilled person would understand that the sleeve insert should be secured to a workpiece. Claim 1 does not lack novelty over McClure 1. 144.Turning to inventive step, I have identified the person skilled in the art and the common general knowledge above. The principal difference between McClure 1 and claim 1 of the Patent is that just identified: there is no teaching in McClure 1 that the sleeve insert should be removed and thus that the fastener as a unit is removable. It goes further. A key teaching of McClure 1 is to anchor the sleeve to the workpiece so that it cannot rotate in use. In my view the skilled person would not have thought it obvious at the filing date to abandon that central feature of McClure 1 which is necessary for the fastener to work. Claim 1 does not lack inventive step over McClure 1. 145.There was a further argument, of secondary importance. It was that McClure 1 nowhere discloses a restriction in the head end of the first member, as required by integer 1(10) of the claim. Integer 1(11) identifies the function of the restriction: to prevent the removal of the second member from the head end of the first member. Mr Ward pointed to the top of the sleeve insert where there is a flange, marked 100 in Figures 15 and 16. He said that was the head of the sleeve insert/first member. I think that is right, but there is no restriction shown there which prevents removal of the collet body/second member from that head. 146.Little was said about whether it was obvious to add a restriction to the head of the sleeve insert/first member of the Figure 15/16 embodiment of McClure 1. Without assistance on this, it seems to me that it was probably obvious. 147.However, claim 1 lacks neither novelty nor inventive step over McClure 1. The same therefore applies to claims 8, 16 and 18. The E-Nut 148.The E-Nut is a physical embodiment of one of the versions of the fastener disclosed in McClure 1. The sleeve insert/first member of the E-Nut must be glued to the underside of the upper panel or workpiece. The collet body/second member is pushed into the sleeve insert, after which the screw is rotated into the collet. The sleeve insert is not removable. Further, Mr Jack accepted in cross-examination that the E-Nut is not a blind fastener because access is required to the far side of the workpiece in order to install it. 149.The E-Nut adds nothing to McClure 1. McClure 2 150.The invention of McClure 2 is entitled “Radial-Type Temporary Fastener, Components and Tool”. An embodiment is shown in Figure 10 which can be used to describe the invention as a whole: 151.There is a main body 30 (the putative first member), a collet 100 (the putative second member) and a spindle 120 (the putative screw member). The collet has flexible fingers at the end distal from the technician (to the left, as shown). The drive nut 60 is rotated and a tool is used to prevent the main body from rotating. Rotation of the nut causes the spindle to be drawn into the main body (to the right, as shown). The bulbous element at the end of the spindle 126 engages with the fingers, which move radially outwards. The heads of the fingers 112 engage the bottom surface of the bottom workpiece. The top and bottom workpieces (not shown) become clamped between the contact portion of the fingers 114 and the contacting surface of the main body 52. There is no disclosure of unclamping and removing the fastener. 152.McClure 2 was relied on by Airbus solely in relation to inventive step. Kwikbolt focussed attention on integer (12) of claim 1 of the Patent: the screw member engaging with a threaded aperture on the second member. 153.McClure 2 begins by referring to a fastener in the prior art. The specification says: “… it can be seen that this type of temporary fastener includes certain inherent deficiencies. One deficiency is that torque applied to a nut that causes translation of the spreader and stem also causes a torque moment to appear in the tubular body tangs. The result is that the tangs rotate or cause a torque moment to appear in one or more of the parts to which the tool is applied. As a result, both the tangs and the part(s) are subject to deformation.” 154.In other words, the torque moment applied to the nut of the prior art fastener is imposed on the fingers, the fingers therefore tend to rotate, which is liable to damage the fingers and/or the workpiece. McClure 2 offers a solution: “A feature of a temporary fastener tool comprising the invention relates to the isolation of expandable collet fingers or tangs from all forms of torque that may result from application of the tool to the work pieces during use. Another feature of the temporary fastener tool comprising the invention relates to the use of a ball and socket configuration as the interface between a spreading element and the collet fingers or tangs.” 155.Mr McKay pointed out that the collet and its fingers are intentionally isolated from the screw member and the collet is not directly operated by the screw member. Thus, the collet does not slide by rotation of the screw member inside the collet, unlike the fastener of the Patent. Mr McKay was not challenged on this evidence. 156.In cross-examination Mr Jack accepted that McClure 2 was telling the reader to isolate the collet fingers from the torque applied to the screw member so that damage would not be caused. He agreed that in contrast the fastener of the Patent requires direct contact between the screw and the collet to open and close the fingers. 157.This idea of isolating the collet fingers from the torque moment of the screw to avoid damaging the fingers and the workpiece seems to me to have been central to the inventive concept of McClure 2. It is also the principal difference between McClure 2 and the Patent. I do not believe that it would have been obvious to a person skilled in the art to abandon the key teaching of McClure 2 and instead have the collet moved by receiving the screw into a threaded aperture within the collet (the second member). Claim 1 does not lack inventive step over McClure 2. 158.Therefore neither claim 8, 16 or 18 lacks inventive step over McClure 2. Wylie 159.Wylie was relied on only as depriving the Patent of inventive step. The invention in Wylie is entitled “Improved Clip for Holding Sheet Metal together”. There are alternative embodiments, all of which have a body (the putative first member), a split-sleeve (the putative second member) and a mandrel (i.e. a spindle). A threaded version of the invention, which includes a putative screw member, is illustrated in Figure 3: 160.The tommy bar 21 is attached to a cap 20, which is attached to a screw 16. A ball-race 22 is located between the cap and the body to reduce friction. There is a split-sleeve 15 in two parts: a wide portion 15b with an internal thread to receive the screw and a narrow portion with a reduced bore at its head 15a. The tommy bar is rotated and with it the screw so that the split sleeve is drawn up into the body 13. The body is prevented from rotating by use of the projecting handle 14. The split-sleeve is prevented from rotating relative to the body by a key and keyway 18 and 19. Attached to the bottom end of the screw there is a mandrel 17. 161.The rotation of the screw causes the split sleeve to move upwards. Its head 15a is drawn over the tapered end of the mandrel, forcing the arms of the splitsleeve to expand radially outwards. The workpieces (not shown) are located between the head of the split-sleeve 15a and the base of the body 13. As the radially expanded arms of the split sleeve move up, they are forced against the lower surface of the lower workpiece. The two workpieces become clamped between the heads of the sleeve arms and the base of the body. 162.The specification explains the release of the fastener (referred to as a “clip”): “In order to release the clip the screw 16 may be unscrewed so as to lift the cap 20 away from the ball race 22 leaving the sleeve in position until the end of the mandrel is clear of the reduced bore of the sleeve thus permitting the head of the sleeve to contract so that the clip may be removed.” 163.The split-sleeve may be split into two parts, so that there are two arms which move apart when the head is pulled over the mandrel. Alternatively it may be in four parts, providing four arms (the putative fingers). Figure 2 shows the latter embodiment in cross-section: 164.By the time of closing argument Kwikbolt had three points, two of them directed to claim 1. The first was based on the cap and the screw moving upwards during removal while the sleeve and the body remain in position. Mr Vanhegan put it this way in closing: “So, it is common ground between the experts that Wylie does not disclose the device being removed as a unit after it has been used to clamp.” 165.That is not at all what either Mr McKay or Mr Jack said. Mr McKay conceded expressly in cross-examination that the fastener was removable as a unit. He accepted that the reference to the sleeve staying in position until the end of the mandrel is clear of the reduced bore of the sleeve meant only that the sleeve cannot be pulled clear of the workpiece until the arms have moved radially inwards sufficiently. 166.At the end of a sustained passage of cross-examination Mr Jack confirmed that the sleeve stays in place until the mandrel is clear of the reduced bore of the sleeve. In my view he was saying the same thing as Mr McKay, which was that once the arms of the sleeve have moved radially inwards, the entirety of the fastener is removed – as a unit. 167.Mr Vanhegan’s second point concerned the same passage of the specification – that quoted above concerning the removal of the fastener. He submitted: “That means that on removing the screw, the second member does not slide relative to the first member whilst simultaneously reducing the radial extent of the finger ends. That is what is required by features 1.15 and 1.16 of the Kwikbolt patent … as accepted by Mr Jack.” 168.Mr McKay did not provide an explanation as to how the Figure 3 fastener could operate unless there were relative motion between the first and second member during removal – in other words, that during removal the sequence of events during insertion is reversed. Mr Jack said that it could not. That makes sense and I accept Mr Jack’s evidence. 169.Before leaving claim 1 I have to consider integer 1(1) – whether Wylie is a removable fastener according to the construction of “removable” that I have found – and integer 1(8) since I have found against Kwikbolt on the construction of that integer. 170.As to integer 1(1), there is no doubt that the Wylie fastener can be removed in the manner discussed in the passage from Wylie I have quoted above. That passage is immediately followed by this: “In such a clip considerable friction takes place between the sleeve and the mandrel and it is advantageous to form both of very hard metal in order to minimise the wear that would otherwise take place.” 171.In his first expert report Mr McKay said that he would expect the friction identified in that passage of Wylie to be caused by the use of a sleeve that has a small clearance in the hole, as described with reference to Figure 1. When discussing Figure 1 Wylie says: “The hole in the articles [the workpieces] is of a size just to permit the passage of the head of the contracted sleeve.” 172.Mr Jack agreed in cross-examination that in the passage of Wylie I have just referred to the reader is being told that a lot of force is required to force the metal of the split sleeve out of the panel members. 173.Wylie contains no clear and unmistakable direction that the fastener is removable in the sense that it may be removed without damage to the fastener or the workpieces. However, Wylie does not suggest that the tight fit of the split sleeve in the apertures of the workpieces is of significance to the invention described. Nor did either expert. It seems to me that it would have been obvious to the skilled person at the priority date of the Patent that the Wylie fastener could be used with workpieces having an aperture of a diameter such that the split sleeve fits less tightly so that the fastener could be removed without damage to the fastener or workpieces. 174.I turn to integer 1(8). This seems to have become lost in the experts’ analysis of Wylie. On Kwikbolt’s construction of integer 1(8) the screw head (the cap and tommy bar) is indeed adjacent the first member head. In consequence, Kwikbolt conceded that this integer was present in Wylie and matters seem to have been left there by both experts. But this concession rests on the head of the first member being at the top of the first member and thus adjacent the cap. I have found that the head of the first member is whichever end it is that bears against the top workpiece. As represented in Figure 3, that is the bottom of the first member. It follows that the head of the first member is not adjacent the screw head, namely the cap and tommy bar. 175.I find it impossible to conceive how Wylie could be made to work with the cap and tommy bar adjacent the end of the first member which bears against the top workpiece. Claim 1 is not obvious over Wylie. 176.Mr Vanhegan’s third argument concerned claim 8. Claim 8 is dependent on claim 1 so it cannot be obvious, but I will address the argument. It was that Figure 3 of Wylie does not disclose planar finger heads. Claim 8 requires that each finger head provides a planar surface to engage with a planar surface of a workpiece to prevent damage to a workpiece. Mr Vanhegan conceded that the Figure 1 embodiment of Wylie was shown as having planar finger heads but submitted that it was not obvious to introduce these into the Figure 3 embodiment. He supported this by evidence from Mr Jack that Wylie was an old disclosure, multi-component and expensive and unlikely to be suitable for automation in use, which was the focus of attention at the filing date. 177.Even if a piece of prior art is old, the hypothesis is still that the skilled person will read it with care at the filing or priority date. If an obvious modification to the prior art would occur to the skilled person, such a modification is not taken out of consideration for inventive step solely because the skilled person would believe that modified or not, this old piece of prior art would have little practical or commercial use as of the filing or priority date. 178.Mr McKay accepted in cross-examination that at the filing date, if the skilled person wanted to use the Figure 3 embodiment for composites, he or she would use flat fingers to spread the load so as to avoid damage. I agree with Mr McKay that taken alone this would have been an obvious modification for fastening materials that were not hard enough to resist distortion.