Prof David Taylor's research reveals that limpets can repair their damaged shells so that they are as strong as the originals. They are still vulnerable to multiple impacts and 'spalling', however - a cause of failure in engineering materials such as concrete

As an academic, most of my research is carried out by supervising PhD students and other researchers. Every now and then, however, I like to get back into the laboratory and do some experiments myself. Last summer, my ‘lab’ was the seashore near my home in Dún Laoghaire and my specimens were limpets. Walking on the beach at Sandycove with my four-year-old niece, collecting shells (as you do), I noticed that some of them had holes in them. And, in a very few cases, there was evidence of a hole having been repaired (see Figure 1).

CLICK TO ENLARGE Figure 1: (a) A typical limpet shell; (b) naturally-occurring damage; (c) damage which has been repaired; (d) damage created by impact testing

Looking into the literature, I found that a major cause of death for limpets and other marine organisms is fracture of shell by impact due to rocks moving around during storms. But, apart from one paper published in the 1980s [1], no mechanical testing has been carried out to investigate the shell’s impact resistance.

So, I made myself a simple impact tester: a metal tube with a series of holes in the side, which allowed me to drop a weight from different heights, giving impact energies from 0.03-1J. I found that by dropping the weight, so that it struck the apex of the shell, I could create damage in the form of a hole that was physically very similar to naturally-occurring damage (Figure 1).

The energy required to do this depended on the size of the shell. Previous workers had shown that limpet shells grow with geometric similarity: all their dimensions, including thickness, increase in linear proportion [2], so the size of the shell can be uniquely characterised by measuring just one dimension.

I used the length L, defined as a the distance across the major axis at the rim (the shape being not quite circular but slightly elliptical). As Figure 2 shows, my results for impact strength as a function of size correlated well with the only previous study, which gave me some confidence in my simple apparatus.

CLICK TO ENLARGE Figure 2: Impact strength (energy to failure) as a function of size for various species of limpet (from [1]) compared with present results for shells containing living limpets tested in situ and empty shells taken from the beach

One might expect that the energy to failure would scale a L3, since (given geometric similarity) this is proportional to the volume of material. Figure 2 shows that this is more or less true across species (except for the very small shells) but, for my particular data, a much better fit was obtained using L4.6. This may be because in my case, damage was always created at the apex, which is thicker than elsewhere (see Figure 3) and may scale differently with length.

Having established a workable experiment, I proceeded to challenge some myths about limpets that one can read in the published literature.

## Physiology of the limpet

CLICK TO ENLARGE Figure 3: (a) Cross section through a limpet shell after impact. Note the thicker region near the apex (b), (c) Scanning electron microscope photos of part of the cross section shown in (a), showing internal damage in the form of delaminations between crystal layers

By way of background, I should mention that the limpet is a kind of snail. It crawls around on the rocks, eating by scraping up plant material using tiny teeth which are extremely hard – one of the hardest materials in nature [3]. Though it moves around, it makes its home in a particular place to which which it always returns.

The rim of the shell is rather irregular in shape because it grows to fit tightly to the rock at this home location. This allows the limpet to minimise water loss when exposed at low tide, and also helps it keep a strong grip when threatened by a predator or a researcher.

The first myth I was able to correct was that this tight fit also helped its impact resistance: I moved limpets to a different location and found that they were just as strong as when at home. The second myth was that the presence of the living animal gave improved impact resistance, either because of its grip on the rock or because its body absorbed some of the impact energy.

I found that empty shells were just as good. However, when I tested shells picked up from the beach they were considerably weaker, requiring only 28% of the energy of fresh shells. After considering several possibilities, I concluded that the reason for this was loss of material at the apex, caused by sand abrasion. Abrading fresh shells with a file until they resembled shells from the beach lowered their impact strength by the same amount.

From my work in other areas – industrial design and forensics – I know that the most common type of failure of engineering components is fatigue, caused by repeated loadings. So I wondered how these limpets would respond to repeated impacts. Very badly, as it turned out. As Figure 4 shows, the energy needed to cause failure after a number of impacts N was, on average, equal to the single-impact energy divided by N. That is to say, failure occurred at a given total amount of energy, whether that was delivered in single blow or over many small impacts.

Compared to the performance of engineering materials, this shows very poor impact resistance: normally, a material can take a much larger number of small impacts or low-amplitude stress cycles (for example, see [4]).

This surprised me, because previous work on other kinds of sea shells had suggested that they have remarkably good toughness, considering that they are essentially ceramic materials made from almost-pure calcium carbonate. In recent years, a lot of work has been done on nacre, which is found in the abalone and other shells and often seen in jewellery where it is called ‘mother of pearl’.

This material has been shown to have remarkably high fracture toughness, which is attributed to a micron-scale structure of bricks or tiles which act to divert cracks and consume energy by shearing [5]. There is much interest in creating bio-inspired engineering materials using the same type of microstructure.

## Fatigue resistance and laminated structure

One would expect that a material with such excellent resistance to crack growth would be able to sustain a large amount of microscopic damage before failing, giving it good fatigue resistance. I believe the explanation for this anomaly lies in the laminated structure.

The shell of the limpet is built up from layers, rather like an onion. The different layers have different types and orientations of calcium carbonate crystals in the calcite or aragonite forms. This creates a shell that is very resistant to the growth of through-thickness cracks, but the strength of the interface between one layer and another is weak, making the structure susceptible to failure by delamination.

CLICK TO ENLARGE Figure 4: Results from impact fatigue tests, plotting the number of impacts to failure as a function of normalised impact energy

I found that, after a non-critical impact, internal delaminations occurred (see Figure 3). Pieces of shell were found to spall off at the apex, on the outer surface where the impactor made contact, but also on the inner surface opposite the contact point. The impact will create a complex stress field, different from that caused by static loading, with significant elements of shear and tension across these interfaces. Similar kinds of delamination failure are common in engineering materials made by layering, especially fibre composite laminates [4].

As I mentioned above, limpets can repair damage to their shells. They cannot actually replace the damaged area, but they can add extra material beneath it to make a kind of patch. My experiments to investigate this phenomenon were not very successful. I put holes in the shells of 33 limpets and left them alive on the rocks to see how well they repaired the holes. But I found that all but two of them disappeared within four weeks.

I am not sure what caused this: possibly gulls or other predators noticed the holes and took advantage of them. Another explanation is increased dehydration or overheating during hot summer days. So this is an experiment which needs improving, but the good news is that when I retested the two surviving limpets, I found that their repaired shells were stronger than the original undamaged shells had been. I hope to return to the rocks this summer to investigate this aspect in more detail.

For those interested in knowing more, the research described in this article has been published in full in the Journal of Experimental Biology [6].

Author: David Taylor, Trinity Centre for Bioengineering, Trinity College Dublin.
Taylor, a chartered engineer, is professor of materials engineering at Trinity College Dublin. His work on the strength and fracture of materials covers biomedical engineering, forensics and industrial design. He is a member of the Royal Irish Academy and a fellow of Engineers Ireland.

References:
[1] Shanks, A.L. and Wright, W.G. (1986) ‘Adding teeth to wave action: the destructive effects of wave-borne rocks on intertidal organisms.’ Oecologia 69, 420-428.
[2] Cabral, J.P. and Natal Jorge, R.M. (2007) ‘Compressibility and shell failure in the European Atlantic Patella limpets.’ Marine Biology 150, 585-597.
[3] Barber, A.H., Lu, D. and Pugno, N.M. (2015) ‘Extreme strength observed in limpet teeth.’ J.Royal Soc. Interface 12, 20141326.
[4] Schrauwen, B. and Peijs, T. (2002) ‘Influence of matrix ductility and fibre architecture on the repeated impact response of glass-fibre-reinforced laminated composites.’ Applied Composite Materials 9, 331-352.
[5] Barthelat, F., Tang, H., Zavattieri, P.D., Li, C.M., and Espinosa, H.D. (2007) ‘On the mechanics of mother-of-pearl: A key feature in the material hierarchical structure.’ Journal of the Mechanics and Physics of Solids 55, 306-337.
[6] Taylor, D. (2017) ‘Impact damage and repair in the shells of the limpet Patella vulgata.’ J.Exp Biol 219, 3927-3935