A day in the life of a Metallurgy Expert Witness

EROS INTRIGUES

Introduction

As part of my work in the early 1990's I was involved with the repair of the Shaftesbury Memorial statue, ‘Eros’, after it was damaged by a football fan who climbed up and swung on the outstretched rear leg. This action caused a fracture in the supporting leg which needed to be addressed. When working with such an important historic monument any action taken must be appropriate for all aspects of the structure, including technical and historical. For this reason it was necessary to find out more about the background to the stature and how it was made.

Background Information

The statue forms part of the memorial to Anthony Ashley Cooper the 7th Earl of Shaftesbury a well-known philanthropist and factory reformer. The memorial was commissioned in the late 1880's at a time when other commemorative works tended to be bronze representations of the person creating dark austere figures. Originally Sir Joseph Boehm was commissioned for the memorial and it was only when he was turned it down that Alfred Gilbert RA was chosen and he decided to go down a very different route.

Gilbert had been educated at Paris’s Ecole-des-Beaux-Arts and specialised in nudes and for this work he chose to represent the good works of Lord Shaftesbury in the form ‘Anteros’, the God of Selfless Charity. He also chose to use aluminium a novel material at that time which was still regarded as semi-precious, like silver. It was only the recent development of the Deville and Castner processes, at that time, which enabled production of aluminium in sufficient quantity to make the world’s first major aluminium sculpture possible.

The statue consists of 15 castings produced in plaster of Paris moulds with mechanical joints that were pinned together and sealed to prevent water ingress. It was made from aluminium that was as pure as it could be produced at the time, approximately 98.5 % aluminium. It is this high level of purity which accounts for its excellent condition even though it has been exposed to the weather and pollution in central London since 1893.

In line with his training Gilbert produced ‘Anteros’ as a naked figure, based on his 16 year old apprentice, Angelo Colarossi. This winged figure appeared to overbalance forward with no external support and this was only possible by the use of this novel lighter aluminium.

When the memorial went on display in 1893 in the centre of Piccadilly it caused quite a stir. The presence of a winged, bright metal, nude figure on top of the ornate bronze fountain was immediately associated with ‘Eros’, the God of love and passion, and it is by this name that the statue is more commonly known.

Despite the subsequent success of the Eros statue this work was the start of Gilbert’s financial problems. The value of the commission was for £3000 but when it was complete it had cost Gilbert £7000.

A little known fact is that Gilbert produced a further five of these figures at which were sold privately and never put on public display. Later, in the 1930's, Gilbert was commissioned to produce a second figure for public display which was mounted on a fountain at Sefton Park in Liverpool. This second figure was made using the same moulds but a different stronger aluminium alloy, containing significant additions of copper and zinc. Unfortunately this higher strength alloy did not have same high corrosion resistance and it suffered from excessive corrosion. Eventually the statue was removed for restoration after which it was displayed indoors to prevent further damage.

Over the life of the Eros statue has been taken down a several times, once in the 1930’s, when it was damaged and an attempt was made to strengthen the support leg by filling it with molten aluminium. It was also removed during the Second World War to prevent possible damage during the bombing and then again in the 1980’s for refurbishment. It may be that these removals and replacements gave rise to the ‘urban myth’ that the statue had been replaced facing the wrong way. It has been suggested that as a memorial to Lord Shaftesbury the statue should not be facing down Piccadilly, with its back towards Shaftesbury Avenue. However, examination of early photographs showing the statue, including one shown in the Magazine of Art, Issue 16, 1892-1893, confirm that the statue has always faced along Piccadilly.

Figure 1 Statue of Eros taken from Magazine of Art

Being such a well-known statue in a prominent place Eros has been subject to various acts of vandalism over its life. The most severe being the damage in 1992 just before its centenary when a football fan swung on the outstretched rear leg and cracked the support leg. It was following this damage that I became involved in the technical aspects of the restoration work.

Investigation of Damage

The first stage was to investigate the extent of the damage. An examination of the statue still in situ found that the crack started at a large aluminium plug in the calf and extended for more than half of the circumference, as shown in Figure 2 below. Due to the extensive damage it was decided to remove the statue for further investigation.

In the workshop the statue was examined non-destructively using liquid penetrant inspection, LPI, and X-rays. The LPI examination revealed many small dowels and casting defects associated with the crack.

Figure 2 Crack passing around calf of support leg

The LPI also revealed a defect on the front of the thigh of the support leg, Figure 3. This was unrelated to the damage caused by the football fan and appears to be related to corrosion under a very thin skin of aluminium over the joint.

Figure 3 Burst at top of thigh opened to reveal it was a thin surface layer of the casting that been disrupted

The X-ray examination revealed the presence of a fill material in the support leg, Figure 4. Further research confirmed that in the 1930's an attempt had been made to reinforce the statue by pouring molten aluminium into the support leg. This attempted reinforcement may account for the larger plug on the rear of the calf which did not appear to relate to the mechanical support of manufacture of the statue. To the top of the X-ray is the solid section where the spigot connection is formed between the leg and body and where the armature is firmly attached to the aluminium.

Figure 4 X-ray through support leg showing iron armature and iron rod core supports for casting

To investigate the quality of the aluminium in the damaged area a core sample was removed through the crack in the support leg. This sample was initially used to examine the fracture faces. The crack surfaces were coarse and uneven, with no indication of fatigue. It was apparent that there was no bond between the aluminium fill material inside the leg and the original leg casting. This was due to the core debris from casting on the inside of the leg. Due to oxidation of the crack faces there was little more that could gained from this examination.

The sections were then mounted and prepared for examination under the microscope. Examination of the cast material confirmed it was primarily aluminium and that there were a large number of defects present. These defects included porosity due to shrinkage as the metal solidified and oxide films which are difficult to avoid when casting aluminium. These are typical problems when casting thin sections, such as those used to make the Eros statue.

Similar casting defects were found in the aluminium fill material in the leg. When combined with the X-ray results it would appear that this material had simply balled up without bonding to the inside of the leg casting and hence provided little or no increase in strength.

The X-ray examination revealed that the iron armature, which sticks out from the bottom of the support leg, passes up though the leg to the structural joint where the leg is attached to the body. This armature had probably been used to support the statue when it was made. Also shown are the steel bars inserted into the cores, to provide strength during casting. Although the cores need to be broken and removed after casting it would not have been possible to remove these core supports without considerable effort and possible damage to the casting, therefore these had been left in-situ.

Possible Remedial Options

As an important piece of cultural heritage the statue attracts a lot of visitors. However, if it was to be left in place it required a structural repair, which to be successful would need to address both conservation and technical issues. Options considered were:

Option 1

Erect a protective barrier over and around the statue keeping the weather off it and at the same time preventing public access. Such a barrier would require regular cleaning/maintenance and would be an obstacle between the public and the statue. This was considered unacceptable for long term service. However, it is understood that this may provide a means of temporary protection, such as turning Eros into a ‘Snow globe’ for the Christmas period.

Option 2

Remove the statue and exhibit it internally putting a replica in its place. Although feasible, since the original moulds are still available at the Victoria and Albert museum, this would mean that the original statue which people come to see would not be on display in Piccadilly. This option was also considered unacceptable.

Option 3

Restore the statue such that it was mechanically sound and could be replaced in Piccadilly. At the same time investigate possible means to make it less susceptible to damage. This was the preferred option as it meant that Gilbert’s original statue would still be available to the public.

Remedial Works

Due to the extent and width of the crack, plus the large number of casting defects and small dowels, etc., around it, the repair required removal of a band of damaged material. Since the crack extended around more than two thirds of the leg circumference, the replacement aluminium castings had to be cast in two pieces. These sections were cast in aluminium with a similar composition to the original material and using original moulds from the Victoria and Albert Museum to achieve the required profile.

The sections were inserted into the leg and then welded into position using Tungsten Inert Gas, TIG, welding which provides greater control to produce a fine repair weld. The welding consumable was also matched to the original composition. After this repair the statue was tested using LPI and unfortunately defects were found along the edges castings in the calf and around the repair welds in the thigh, Figure 5. These defects were due to a combination of castings defects in the original material and shrinkage due to the composition of the weld consumable.

Figure 5 Defects opening around the weld after initial repair

For this reason the original repairs were cut out and the joints remade. For the second repair the statue was insulated and preheat applied to reduce the cooling rate of the welds. Also an aluminium silicon weld consumable, with both lower melting point and lower shrinkage characteristics, was used.

Dye penetrant testing of the re-welded repair confirmed that the welds were sound. The profiles of the repairs were then worked by a sculptor to match the original profile, after which it was retested and confirmed to be sound.

Return of the Statue

When the statue was returned to Piccadilly changes to the way in which it was mounted were made. Instead of bolting through the aluminium plate around the foot of the support leg the top of the fountain was modified to form a socket for the iron armature. Supporting the statue by the iron armature

reduces the stress in the aluminium leg castings making it more resilient to possible damage in the future.

Once the statue had been positioned on the fountain the armature was secured in position and the underside of the aluminium was sealed to the fountain completing the repair, as shown in Figure 6.

Figure 6 Supporting leg of repaired statue when returned to Piccadilly showing sealed mounting holes where statue is now mounted on the iron armature

Eur Ing SIMON CLARKE

BSc(Hons) MBA CEng CSci MIMMM FInstNDT

MWeldI MICorr ACQI FIAQP MAE MEWI