Posted to News on 26th Mar 2012, 00:00
A decade of vibration testing on inserts: the results are in
In 2007, we reported on the surprising locking ability of the Spiralock threaded insert. Now a decade of testing has revealed that it actually works much better than was originally thought.
>A report written in 1986 by David Light of British Aerospace, Naval Weapons Division concluded that the Spiralock wire thread insert manufactured by the Cross Manufacturing Company was an excellent bolt locking system when subjected to various rigorous tests of transverse vibration. This conclusion was reached in comparison to another commercially available screw locking insert. This raised a question at Cross Manufacturing: exactly how good was it?
>To answer the question, in 2001 the company began a programme of testing at its Bath premise. The purpose of this programme was to assess in detail how well the insert works, especially in relation to the full range of alternative commercially available locking systems. Now, some ten years and 1500 short and long-term tests later, the true effectiveness of this locking device can be fully revealed.
>The Spiralock wire thread insert is manufactured from an 18/8 stainless steel (DTD 734A) and is based on a 60° thread form, but includes a 30° ramp which is the bearing surface for the bolt thread. During assembly into the insert, a bolt will require some 10 to 20% higher torque in the Spiralock insert than in a standard 60° thread form in order to achieve the same preload. Tests have shown that this extra torque creates increased radial forces. The resulting combination of tensile and radial forces ensures that the bolt, the insert and the material into which it is inserted (parent material) are securely locked together.
>A recommended torque of 75%, 0.2% proof stress for an 8.8 property class bolt, using lubrication of two drops of light engine oil on the bolt thread, produces consistently secure bolt locking results. Tests have revealed that under head lubrication does not reduce the frictional effect of the assembly nor significantly improve the locking efficiency. The required length of insert depends on the shear strength of the parent material. For example, light alloys, steels and titanium alloys should use an insert whose length is one and a half times the nominal diameter of the bolt. For weaker magnesium alloys the figure for the insert length is twice the nominal bolt diameter dimension and for the much lower strengths of plastics tested the figure is three times. Surprisingly, it has been shown that increases in these recommended lengths of engagement do not require different torque values to achieve secure locking as the locking efficiency is not affected.
>Using the recommended tapped hole, insert size, lubrication and torque value ensures secure bolt locking in all of the materials tested, whether metals or plastics. So just how was the Spiralock inserted tested? The inserts were made in both Metric and Unified thread formats. In most cases, the inserts were unplated, but for specific applications silver plated inserts or IVD aluminium coatings were used. A wide range of both metallic and plastic parent materials were tested. The metals included light alloy, magnesium alloy, steels and titanium alloy whilst the plastics included Sustarin C, Nylon 6, GRP and PET. More than 40 different lubricants were investigated.
>Testing process
>The tests included black bolts (8.8 and 12.9 property classes), some plated with cadmium or zinc. High strength bi-hex aerospace bolts and washers in Inconel 718 were widely used to replicate aerospace applications. In addition, stainless steel bolts (specifications A2 70 and A4 80) were tested. All the tests were carried out on a Junkers transverse vibration rig. This enabled accurate test information to be continuously and automatically recorded using a digital data acquisition system.
>The tests included: simple load/unload cycles with no vibration to assess frictional effects of the assemblies; basic vibration tests for 6500 cycles to check the locking efficiency; basic vibration tests repeated 100 times using the same bolt/insert installation; continuous vibration testing for over 20 hours to complete one million cycles; investigation of minimum preload locking force required to ensure efficient bolt locking; multiple six hour heat soaks at 650°C followed by vibration testing, with and without disassembly between testing cycles; and pullout tests.
>One of the most enlightening results of the test programme was the reduced preload cycle. Having tightened the bolt to the recommended torque value to achieve the design preload, the bolt was partially loosened, thus reducing the preload. A 'latching' effect was then apparent, with secure locking remaining until the bolt preload was reduced to approximately 20% of 0.2% proof stress.
>Pullout test results were also interesting. Whilst pullout on any of the metallic parent materials was never observed (the selected bolt was designed to fail first), pullout from some plastics did occur. Whilst these specimens efficiently locked at the required preload value, excessively high preloads caused pullout to occur. The pullout value was greatly in excess of the recommended value (up to 25% higher) and confirmed the increased benefit of the radial forces generated when the bolt is torqued in the Spiralock insert.
>Comparative testing against a wide range of commercially available head and thread locking systems investigated the relative locking efficiency of the Spiralock insert. Comparative systems included various threads, washers, wire thread insert, thin wall inserts and various adhesives. Initially, each pair of locking systems was tested at low vibration amplitude. If both systems successfully remained locked, another test was applied at increased amplitude. This process continued until one of the types failed, by either bolt loosening or fatigue.
>All of the systems worked well at the basic amplitudes but all ultimately failed except for the Spiralock wire thread insert. It was surprising to observe that during the 650°C thermal testing, a comparison with a silver plated wire thread screw locking insert made from a high temperature alloy resulted in failure after a single cycle whilst the stainless steel unplated Spiralock insert was still locking efficiently after six full cycles of the test.
>The success of the insert is basically reliant on its manufacture. The18/8 stainless steel wire is rolled to the required cross-section which is critically dimensionally assessed, with all design angles maintained to a tolerance of ±0.5 degrees. The wire is then coiled and the coils finished by hand, including cutting to length, creating the tang and the tang removal notch. Rigorous inspection is applied at each operation.
>There are additional benefits associated with the use of the Spiralock thread insert. Firstly, the installation of the insert into the tapped hole is very easy (no pre-winding is required) and since there is no prevailing torque, the bolt is free running and can be screwed into the insert by hand. For maintenance applications, once the preload is fully removed the bolt can be unscrewed by hand (no prevailing torque). Multiple disassemblies and reassemblies can be carried out with no loss of locking performance. In addition the insert can be easily removed from the tapped hole using a simple extraction tool with no damage to the internal threads.
