ZU-DBV, ZU-RHO & RAF 2000 hub bar bolt torque Threads Merged
Posted: Sat Apr 29, 2017 8:12 pm
MODERATED
The more I thought about the ZU-RHO accident report I realised that the AR failed to identify the root cause of the accident. Refer to my attached formal objection to the SACAA. Whether the SACAA is going to act on it or not, is not important to me. I notified them and my conscience is clear.
The PROBLEM, however, remains unresolved. By problem I mean the bolted joint in question. First of all, the RAF NAS 501-012 AFC bolt used in the assembly is a custom made bolt produced for RAF exclusively. (I could gather no information to indicate the contrary.) The properties and characteristics of any other AN or NAS bolt used in the aerospace industry is readily available in the public domain and production thereof is subject to rigorous quality and conformance control, but the RAF NAS 501-012 AFC bolt on which I must trust my life, is cloaked in darkness. Rafsa blankly refuses to supply the information so it is also possible that it officially does not exist. There is also the problem with the chamfered washer used under the bolt head in addition to the complete lack of a procedure describing the modification process thereof. PN-39 also does not describe the washer used under the nut. Crucial to the proper operation of this tension joint is the procedure of assembly and quality control thereof which, sadly, is also cloaked in darkness. Do not think that the SACAA is going to cause the problem to disappear, it is not their beef, it is the beef of the owner/operator of the ZU craft.
I rest my case for now.
Rgds
Francois Marais
My Objection :
DATE : 10 APRIL 2017
TO : The Director SACAA and
Mr J Visser : SENIOR AIRCRAFT ACCIDENT AND
INCIDENT INVESTIGATOR SACAA
FROM : HF MARAIS 0272201179 ZU-DTS
ID 5810065081082
PO Box 585
GANSBAAI, 7220
0827712073
FORMAL OBJECTION TO ACCIDENT REPORT CA18/2/3/9462 RE ZU-RHO
Sir,
Herewith my formal objection re the content and findings of the AR above. The withheld and/or misinformation presented to the SACAA skewed the investigation and therefore the SACAA was prevented from reaching a definitive conclusion. Refer Appendix D Para [12]. I have no objection to the CrashLab report. The bolt in question failed simply because a non conforming washer was installed under the bolt head, implying that the proper procedure was not followed and the subsequent quality control failed or was non existent. Negligence was the cause of this accident which very easily could have cost the lives of a father and a son. I will endeavour to explain.
The information as per AR CA18/2/3/9462 Para 1.16.7 : RAFSA Procedure No7 issued April 2007, revised 28 February 2011 (rev 001) which was supplied by RAFSA to the SACAA is incorrect and incomplete. This assembly procedure, including pictures, relates to the use of the AN12-34A bolt, NOT the RAF NAS 501-012 AFC bolt, washers and castellated nut assembly which failed. Please compare the photographs of this paragraph to that of the CrashLab report. (Para 1.16.8.1 is another point in case - apart from the grossly erroneous stress level figures quoted for the bolt.)
RAF Product Notice – 39, dated 15 December 2004, announced the introduction of the RAF NAS 501-012 AFC bolt, washers, new form winglet and castellated nut. I see that an incomplete copy of this PN-39 is attached to the AR as appendix F. I do not know if this was presented to SACAA by RAFSA as published, but I would like to point out that it is also incomplete as it does not show the pictures of the mentioned hardware. For clarity I will refer to the RAF NAS 501-012 AFC assembly currently in use as the ‘new assembly’ and to the superseded AN12-34A assembly as the ‘old assembly’. The ‘new assembly’ incorporates a chamfered and modified washer under the bolt head as well as a washer and castellated locknut with split pin on the other end of the assembly. The modified washer is a specially chamfered hardened, round, flat washer of which two opposite (round) ends are trimmed (cold ground or cold sawed) straight according to the bolt head dimension in order for it to not contact the rounded surface of the hub bar when the bolt is fixed. CrashLab photo no 12 shows these washers and evidently they were totally round and not modified. The ‘old assembly’ did not have a washer under the bolt head and utilised an uncastellated locknut and washer. Due to these fundamental differences, the assembly procedures for these two vastly different configurations can NOT be the same.
As background I present the following information. The kit for my RAF was ordered somewhere in 2004 and completed early 2005. It came with the old style winglets and the ‘old assembly’ hub bar. In the beginning of 2006 another pilot had a roll-over with this gyro and amongst the replacement spares, new rotor blades and hub bar was ordered. What arrived from RAF Canada was the new generation winglet with the ‘new assembly’ hub bar. So this was beginning 2006. In the beginning of 2008 I noticed a very fine crack developing on the side of the winglet closest to the cylindrical steel insert used as bearing surface for the pitch adjustment screw. I ordered a complete new hub bar from Rafsa as Rafsa informed me that RAF Canada refused my claim. What I received from Rafsa was again the new generation assembly. That was beginning 2008. Both hub bars thus was complete with the ‘new assembly’ incorporating the correctly modified washers and this leads me to believe that a procedure for the ‘new assembly’ as well as a specification for the chamfered washer and it’s modification had to exist from early on. Especially since RAF Canada was subject to the American law system.
Now the question is, why would the assembly procedure, RAFSA “Procedure No7 issued April 2007”, describing the ‘old assembly’, be issued in April 2007 if the hardware production and assembly had been announced to change December 2004 and there is evidence that production had indeed changed by beginning 2006? Even more astounding is the fact that this procedure was revised, still wrt the outdated ‘old assembly’, in February 2011! : RAFSA “Procedure No7 issued April 2007, revised 28 February 2011 (rev 001)” as per this AR. Very strange indeed.
Furthermore, according to Appendix C : SACAA and RAFSA Meeting on 21 September 2015 point 7, it is stated “The organisation was unable to provide the design data, not even in the form of a drawing of the washer used with the effected bolt during the visit. They stated that the washer was deemed a standard item.” I find this ignorance very concerning. I can only assume that Rafsa at this point did not inform SACAA that the washer needs to be modified for use in the assembly. This washer on it’s own is also a critical part of the joint and must be of hardened steel with adequate properties in order not to compromise the joint. Inspection of this modified washer showed me that the modification is a procedure performed after it was produced and plated in the factory. This modification process needs to be specified and quality controlled so that the washer does not heat up sufficiently to lose it’s temper during this process. Clearly, absence of an approved procedure is a grave shortcoming. I (and others) have previously indicated to Rafsa that I am not happy with the grossly excessive chamfer to this washer. This arrangement unduly restricts the available area for load transfer between the bolt head and the washer and leads to increased stress and bending moment levels to the head of the bolt. In due course I will endeavour to quantify my argument. This washer is not just ‘a standard item’.
Rafsa publicly (it is even quoted in AR) claims safety factors of 4 to 5 for the bolt in question. This is not true. My initial calculations assuming reasonable bolt parameters and considering ONLY the main forces at play, indicates a SF of 1.8 to 1.9, but this could change either way depending on how the joint is assembled. A split pin can only be inserted at 60 degree turning intervals on the castellated nut which means that you can only achieve specific torque intervals when fixing the joint. These ranges are also going to differ from joint to joint if there is significant dimensional variation in the components of the joint. Temperature is also going to have it’s hand in the pie. Obviously in this arrangement the position of the torque ranges can be altered by the use of shims, but the 8 joints which I had a look at, contained no shims. All these things make this critical joint a complicated one. Experiments that I have done since the AR was released, indicated that on this joint there are only three slots available to turn the nut to after the ‘snug’ position. Either by design, or by coincidence, the hole in the bolt lined up more or less with a slot in the nut at the snug position. Torque on the joint at the first slot will be fatally low, at the second slot is seems to be in the right region and at the third slot it is too high as deformation of the aluminium under the washer starts taking place. I have to mention here that it is a recognised practice in industry to torque a joint by tightening the nut a certain number of degrees from the snug position and this can, under controlled conditions, be more accurate than using a torque wrench which anyway, even if calibrated correctly, yields uncertainty of +/- 25%. I am compelled to come to the conclusion that this is the way that the joint under discussion is being assembled.
I can now only wonder what the person who assembled RHO’s hub bar did. By turning the unmodified washer so that it’s unmodified sides touches the rounded sides of the hub bar, as would be the case as seen in RHO’s hub bar, a gap of about 1,5mm exists between the washer and the hub bar where it must seat. That equates to force turning the nut (and in the process deforming the washer) through at least 6 consecutive slots before the washer seats against the hub bar. It will now only be in the snug position offering no stiffness to the joint. The nut will still need to be turned through two more slots for the joint to be able to work. So what the operator did now is anyone’s guess, but whatever he did from here, one slot or two slots further, the failure mechanisms had already been set up and sooner or later the end result would have been the same – the bolt did not have a chance. From this we learn a number of things about the operator who assembled RHO’s hub-bar. The most important being that he(she) had no technical background, had no training with respect to the procedure he was about to perform and had no previous experience assembling the RAF hub bar. Had he been simply told to snug the washer against the hub bar and then turn two more slots? This also tells a story about the facility. Could this have happened in Canada or did this happen at Uppington? Whatever the case, gross negligence was perpetrated and accountability in this case is clear.
From practice we know the joint works but we do not know how well it works. I must add that it was a great relief for me to realize that RHO’s bolt failed due to negligence and not anything else. For me, or anyone for that matter, to analyse what we are dealing with here, it is necessary to know the specifications for the RAF NAS 501-012 AFC bolt, washers, nut and the assembly procedure. I have on a previous occasion asked Rafsa kindly to provide me with these details for purpose of analysis, but was stonewalled completely, them claiming proprietary confidentiality. This is utter nonsense. Information on any approved fastener product used in the aviation industry is available in the public domain for reason of accountability (amongst others) and Rafsa can be no exception to this rule. Even the more so in this critical application where custom made fastener equipment is used. In the interest of safety and accountability, this information must be included in the AR.
In view of the above, to be fair to the survivors of ZU-RHO, to serve the general aviation community as a whole and in the spirit of truth, transparency and safety, I kindly request you to re-open the investigation and revise AR CA18/2/3/9462.
Thanking You &
Kind Regards
_____________
Francois Marais
The more I thought about the ZU-RHO accident report I realised that the AR failed to identify the root cause of the accident. Refer to my attached formal objection to the SACAA. Whether the SACAA is going to act on it or not, is not important to me. I notified them and my conscience is clear.
The PROBLEM, however, remains unresolved. By problem I mean the bolted joint in question. First of all, the RAF NAS 501-012 AFC bolt used in the assembly is a custom made bolt produced for RAF exclusively. (I could gather no information to indicate the contrary.) The properties and characteristics of any other AN or NAS bolt used in the aerospace industry is readily available in the public domain and production thereof is subject to rigorous quality and conformance control, but the RAF NAS 501-012 AFC bolt on which I must trust my life, is cloaked in darkness. Rafsa blankly refuses to supply the information so it is also possible that it officially does not exist. There is also the problem with the chamfered washer used under the bolt head in addition to the complete lack of a procedure describing the modification process thereof. PN-39 also does not describe the washer used under the nut. Crucial to the proper operation of this tension joint is the procedure of assembly and quality control thereof which, sadly, is also cloaked in darkness. Do not think that the SACAA is going to cause the problem to disappear, it is not their beef, it is the beef of the owner/operator of the ZU craft.
I rest my case for now.
Rgds
Francois Marais
My Objection :
DATE : 10 APRIL 2017
TO : The Director SACAA and
Mr J Visser : SENIOR AIRCRAFT ACCIDENT AND
INCIDENT INVESTIGATOR SACAA
FROM : HF MARAIS 0272201179 ZU-DTS
ID 5810065081082
PO Box 585
GANSBAAI, 7220
0827712073
FORMAL OBJECTION TO ACCIDENT REPORT CA18/2/3/9462 RE ZU-RHO
Sir,
Herewith my formal objection re the content and findings of the AR above. The withheld and/or misinformation presented to the SACAA skewed the investigation and therefore the SACAA was prevented from reaching a definitive conclusion. Refer Appendix D Para [12]. I have no objection to the CrashLab report. The bolt in question failed simply because a non conforming washer was installed under the bolt head, implying that the proper procedure was not followed and the subsequent quality control failed or was non existent. Negligence was the cause of this accident which very easily could have cost the lives of a father and a son. I will endeavour to explain.
The information as per AR CA18/2/3/9462 Para 1.16.7 : RAFSA Procedure No7 issued April 2007, revised 28 February 2011 (rev 001) which was supplied by RAFSA to the SACAA is incorrect and incomplete. This assembly procedure, including pictures, relates to the use of the AN12-34A bolt, NOT the RAF NAS 501-012 AFC bolt, washers and castellated nut assembly which failed. Please compare the photographs of this paragraph to that of the CrashLab report. (Para 1.16.8.1 is another point in case - apart from the grossly erroneous stress level figures quoted for the bolt.)
RAF Product Notice – 39, dated 15 December 2004, announced the introduction of the RAF NAS 501-012 AFC bolt, washers, new form winglet and castellated nut. I see that an incomplete copy of this PN-39 is attached to the AR as appendix F. I do not know if this was presented to SACAA by RAFSA as published, but I would like to point out that it is also incomplete as it does not show the pictures of the mentioned hardware. For clarity I will refer to the RAF NAS 501-012 AFC assembly currently in use as the ‘new assembly’ and to the superseded AN12-34A assembly as the ‘old assembly’. The ‘new assembly’ incorporates a chamfered and modified washer under the bolt head as well as a washer and castellated locknut with split pin on the other end of the assembly. The modified washer is a specially chamfered hardened, round, flat washer of which two opposite (round) ends are trimmed (cold ground or cold sawed) straight according to the bolt head dimension in order for it to not contact the rounded surface of the hub bar when the bolt is fixed. CrashLab photo no 12 shows these washers and evidently they were totally round and not modified. The ‘old assembly’ did not have a washer under the bolt head and utilised an uncastellated locknut and washer. Due to these fundamental differences, the assembly procedures for these two vastly different configurations can NOT be the same.
As background I present the following information. The kit for my RAF was ordered somewhere in 2004 and completed early 2005. It came with the old style winglets and the ‘old assembly’ hub bar. In the beginning of 2006 another pilot had a roll-over with this gyro and amongst the replacement spares, new rotor blades and hub bar was ordered. What arrived from RAF Canada was the new generation winglet with the ‘new assembly’ hub bar. So this was beginning 2006. In the beginning of 2008 I noticed a very fine crack developing on the side of the winglet closest to the cylindrical steel insert used as bearing surface for the pitch adjustment screw. I ordered a complete new hub bar from Rafsa as Rafsa informed me that RAF Canada refused my claim. What I received from Rafsa was again the new generation assembly. That was beginning 2008. Both hub bars thus was complete with the ‘new assembly’ incorporating the correctly modified washers and this leads me to believe that a procedure for the ‘new assembly’ as well as a specification for the chamfered washer and it’s modification had to exist from early on. Especially since RAF Canada was subject to the American law system.
Now the question is, why would the assembly procedure, RAFSA “Procedure No7 issued April 2007”, describing the ‘old assembly’, be issued in April 2007 if the hardware production and assembly had been announced to change December 2004 and there is evidence that production had indeed changed by beginning 2006? Even more astounding is the fact that this procedure was revised, still wrt the outdated ‘old assembly’, in February 2011! : RAFSA “Procedure No7 issued April 2007, revised 28 February 2011 (rev 001)” as per this AR. Very strange indeed.
Furthermore, according to Appendix C : SACAA and RAFSA Meeting on 21 September 2015 point 7, it is stated “The organisation was unable to provide the design data, not even in the form of a drawing of the washer used with the effected bolt during the visit. They stated that the washer was deemed a standard item.” I find this ignorance very concerning. I can only assume that Rafsa at this point did not inform SACAA that the washer needs to be modified for use in the assembly. This washer on it’s own is also a critical part of the joint and must be of hardened steel with adequate properties in order not to compromise the joint. Inspection of this modified washer showed me that the modification is a procedure performed after it was produced and plated in the factory. This modification process needs to be specified and quality controlled so that the washer does not heat up sufficiently to lose it’s temper during this process. Clearly, absence of an approved procedure is a grave shortcoming. I (and others) have previously indicated to Rafsa that I am not happy with the grossly excessive chamfer to this washer. This arrangement unduly restricts the available area for load transfer between the bolt head and the washer and leads to increased stress and bending moment levels to the head of the bolt. In due course I will endeavour to quantify my argument. This washer is not just ‘a standard item’.
Rafsa publicly (it is even quoted in AR) claims safety factors of 4 to 5 for the bolt in question. This is not true. My initial calculations assuming reasonable bolt parameters and considering ONLY the main forces at play, indicates a SF of 1.8 to 1.9, but this could change either way depending on how the joint is assembled. A split pin can only be inserted at 60 degree turning intervals on the castellated nut which means that you can only achieve specific torque intervals when fixing the joint. These ranges are also going to differ from joint to joint if there is significant dimensional variation in the components of the joint. Temperature is also going to have it’s hand in the pie. Obviously in this arrangement the position of the torque ranges can be altered by the use of shims, but the 8 joints which I had a look at, contained no shims. All these things make this critical joint a complicated one. Experiments that I have done since the AR was released, indicated that on this joint there are only three slots available to turn the nut to after the ‘snug’ position. Either by design, or by coincidence, the hole in the bolt lined up more or less with a slot in the nut at the snug position. Torque on the joint at the first slot will be fatally low, at the second slot is seems to be in the right region and at the third slot it is too high as deformation of the aluminium under the washer starts taking place. I have to mention here that it is a recognised practice in industry to torque a joint by tightening the nut a certain number of degrees from the snug position and this can, under controlled conditions, be more accurate than using a torque wrench which anyway, even if calibrated correctly, yields uncertainty of +/- 25%. I am compelled to come to the conclusion that this is the way that the joint under discussion is being assembled.
I can now only wonder what the person who assembled RHO’s hub bar did. By turning the unmodified washer so that it’s unmodified sides touches the rounded sides of the hub bar, as would be the case as seen in RHO’s hub bar, a gap of about 1,5mm exists between the washer and the hub bar where it must seat. That equates to force turning the nut (and in the process deforming the washer) through at least 6 consecutive slots before the washer seats against the hub bar. It will now only be in the snug position offering no stiffness to the joint. The nut will still need to be turned through two more slots for the joint to be able to work. So what the operator did now is anyone’s guess, but whatever he did from here, one slot or two slots further, the failure mechanisms had already been set up and sooner or later the end result would have been the same – the bolt did not have a chance. From this we learn a number of things about the operator who assembled RHO’s hub-bar. The most important being that he(she) had no technical background, had no training with respect to the procedure he was about to perform and had no previous experience assembling the RAF hub bar. Had he been simply told to snug the washer against the hub bar and then turn two more slots? This also tells a story about the facility. Could this have happened in Canada or did this happen at Uppington? Whatever the case, gross negligence was perpetrated and accountability in this case is clear.
From practice we know the joint works but we do not know how well it works. I must add that it was a great relief for me to realize that RHO’s bolt failed due to negligence and not anything else. For me, or anyone for that matter, to analyse what we are dealing with here, it is necessary to know the specifications for the RAF NAS 501-012 AFC bolt, washers, nut and the assembly procedure. I have on a previous occasion asked Rafsa kindly to provide me with these details for purpose of analysis, but was stonewalled completely, them claiming proprietary confidentiality. This is utter nonsense. Information on any approved fastener product used in the aviation industry is available in the public domain for reason of accountability (amongst others) and Rafsa can be no exception to this rule. Even the more so in this critical application where custom made fastener equipment is used. In the interest of safety and accountability, this information must be included in the AR.
In view of the above, to be fair to the survivors of ZU-RHO, to serve the general aviation community as a whole and in the spirit of truth, transparency and safety, I kindly request you to re-open the investigation and revise AR CA18/2/3/9462.
Thanking You &
Kind Regards
_____________
Francois Marais