IEEE Robotics & Automation Magazine - December 2020 - 49

the healing of macroscopic damage takes time. In a second experiment, the healing efficiency, based on the
recovery of the fracture strain and the fracture stress, is
experimentally measured as a function the healing time.
Samples with a width of 5.5 mm and a thickness of
2-2.5 mm were subjected to stress-strain tensile tests
until they fractured [Figure 3(a)]. As a reference, six
(undamaged) samples were fractured in a stress-strain
test [Figure 4(a)]. These samples failed, on average, at an
approximate strain of 245% and a stress of 0.1 MPa. The
Young's modulus of this material (the slope of the tangent
line in the origin of the stress-strain curve) is 0.12 MPa.
Next, 24 samples were sliced in two with a clean scalpel
blade [Figure 3(b)]. Immediately after the cut, the two ends
were manually brought back into contact. When macroscopic
misalignments are avoided while fitting the fracture surfaces
back together, the instant healing ability of the DPBMFT5000-r0.5 network facilitates the precise merging of the
parts so that the cut is no longer visible when investigated
through optimal microscopy [Figure 3(e)-(g)]. These samples
were left to heal at room temperature for one, three, seven, and
14 days. For each healing time, six samples were fractured in a
stress-strain test [Figure 4(a), 1%.s-1). The mean fracture
stresses and strains are presented in the block diagrams in Figure 4(b). In Figure 4(c), the mean healing efficiencies for the
different healing times (Ht) were calculated by comparing the
fracture strains and stresses with those
measured in the reference experiment:

through time clearly proves the contribution of the re-formation of these reversible links to the healing process at
room temperature.
After 14 days of healing at room temperature, the fracture no longer took place at the location where the cut was
made but, rather, at a location where an imperfection caused
stress concentrations [e.g., a cavity caused by a solvent bubble or a dust particle; Figure 3(d)]. Taking into account the
standard error of the mean presented in the block diagrams
in Figure 4(c) and the fact that the fracture did not happen
at the location of the " scar " of the cut, it can be concluded
that, after 14 days, the cuts are completely healed and that
the initial strength of the samples has been completely
recovered. The presented results were all obtained at 25 °C.
At lower application temperatures, healing takes slightly
longer, while at higher temperatures, the duration of healing
is shortened.
Healing Efficiency as Function of Healing Cycles
The samples that were healed for one, three, and seven days
were fractured in the stress-strain tensile test to evaluate the restoration efficiency, and the generated pieces [Figure 3(b)] were
immediately brought back into contact [Figure 3(c)]. After again
healing for the same duration-one, three, and seven days-the
samples were again fractured in the tensile tests. For all healing
times, the fracture strain and the fracture stress of the second

  h f^Ht h = f fract^Ht h /f fract^not damagedh, (3)
  h v^Ht h = v fract^Ht h /v fract^not damagedh.(4)
Figure 4(a) illustrates that, after
healing occurs at room temperature,
very similar stress-strain characteristics are measured, but failure results
at much lower stresses. Due to slow
reaction kinetics, creating interfacial
DA bonds clearly takes time. After
healing for one day at 25 °C, only
50% of the fracture stress (h v) was
recovered [Figure  4(c)]. Visual in--
spection showed that the fracture
took place at the same location where
the cut was made. The formed fracture surfaces again looked clean and
identical to the picture in Figure 3(b).
The healing efficiencies ( h f and h v )
can be increased by prolonging the
healing time. Indeed, after three,
seven, and 14 days, the fracture stress
recovered by, respectively, 62, 91, and
97%. Although DA reactions are generally considered to be too slow for
room temperature autonomous healing, the increasing failure strength

0.5 mm

(a)

(b)

(c)

(d)

0.5 mm
(e)

0.5 mm
(f)

(g)

Figure 3. Testing autonomous healing at room temperature. (a) Samples with a width of
5.5 mm, a thickness of 2-2.5 mm, and a length of 8-10 mm. (b) The samples are sliced
in two using a scalpel blade. (c) The samples are pressed back together seconds after the
damage was inflicted and placed at room temperature for one, three, seven, or 14 days
before being subjected to a stress-strain tensile test until they fracture. (d) The samples
that were cured for 14 days do not fracture at the location of the initial damage but at a
new location. Microscopic images of (e) the sample prior to damage, (f) the sample after
it is damaged, and (g) and the sample when it is reconnected.

DECEMBER 2020

*

IEEE ROBOTICS & AUTOMATION MAGAZINE

*

49
IEEE Robotics & Automation Magazine - December 2020

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - December 2020
