See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/289663184

Manufacturing of mycelium-based biocomposites

Conference Paper · May 2013

CITATIONS READS
14 1,786

4 authors, including:

Lai Jiang
Prairie View A&M University
17 PUBLICATIONS   50 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Facilitating Industry by Engineering, Road mapping and Science (FIBERS) to Advance U.S. Manufacturing of Composites View project

All content following this page was uploaded by Lai Jiang on 13 July 2016.

The user has requested enhancement of the downloaded file.

 MANUFACTURING OF MYCELIUM-BASED BIOCOMPOSITES
Lai Jiang, Daniel Walczyk, Liam Mooney, Samuel Putney
Center for Automation Technologies and Systems
Rensselaer Polytechnic Institute
110 Eighth Street
Troy, NY 12180
ABSTRACT
Sustainable composites that use renewable materials and provide better end-of-life
options are of great interest to industry. This paper describes an investigative study into
high-production manufacturing approaches for a biocomposite material with these
characteristics consisting of natural fiber reinforcement and agricultural waste cores
bound together by a fungal mycelium matrix that grows in and around everything.
Specific processes investigated include cutting of reinforcement plies (woven jute textile
in this case), impregnating individual plies a temporary glue binder, stacking and forming
laminates or „skins‟, drying and sterilization of formed skins, and assembling all
components that comprise a composite sandwich structure prior to the mycelium growth
phase. Optimal processes are determined according to a number of metrics such as
shortest cycle time, lowest cost, lowest energy consumption, and best product quality.
The final manufacturing processes were selected based on the results of this comparative
study.
Key words: biocomposite, composite manufacturing, preformed skin, cutting,
impregnating, stacking and forming, drying

1. INTRODUCTION
Almost all plastic materials used today are made with little regards to sustainable
end-of-life options. When plastic resins are combined with reinforcement fibers such as
glass or carbon, and low-density cores such as foam or honeycomb to make advanced

1
composite parts and sandwich structures with composite skins, their end-of-life options
are even worse. This situation has led to worldwide concerns over the environmental
issues related to the disposal of non-biodegradable plastic and polymer composite
materials after their intended use. The idea of developing biodegradable materials that
have similar performance to traditional engineering materials, but will undergo a
complete but controlled biodegradation just like any other organic waste, has been of
interest to researchers for many years, although most of the current research and
applications have focused on reinforcements [1-3]. Some examples of prior R&D
include:
 Henry Ford, in the 1930s and 1940s, mixing soy protein with
phenol-formaldehyde resin to produce automobile body parts [4]
 Brother and McKinney reportedly making plastics using soy protein and various
cross-linking agents [5]
 Drzal et al. finding the combination of bio-fibers like kenaf, hemp, flax, jute,
henequen, pineapple leaf fiber and sisal with polymer matrices can be used in
automotive applications [6] and
 Flanigan et al. using natural fibers to replace either part or all of the glass fiber
reinforcement used in the composites in order to increase the bio-content [7].

A unique, grown biocomposite sandwich structure, shown in Figure 1, comprised of

natural fibers as reinforcement, agricultural waste material (e.g., Kenaf pith) as core, and
mushroom mycelium, a fungal network of threadlike cells that are essentially the “roots”
of mushrooms, as matrix, has been developed recently by Ecovative Design, LLC (Green
Island, NY) in collaboration with the Center for Automation Technologies and Systems
(CATS) at Rensselaer Polytechnic Institute. The mycelium digests the surface of damp
agricultural byproducts in 5-7 days without additional energy input to act as a natural,
self-assembling glue [8]. Water is driven off and the mycelium is inactived (killed) by
drying the composite in a conventional convection oven. The main advantages of

2
„mycocomposite‟ materials made using natural fibers such as kenaf, jute, hemp, flax,
bamboo, and sisal over traditional synthetic composites are low cost, low density,
competitive strength, tensile and impact mechanical properties, reduced energy
consumption, the potential for CO2 sequesterization, and perhaps most important of all,
biodegradability [6]. The low-embodied-energy production process used for
mycocomposites makes them cost competitive with synthetics, but the material is
inherently sustainable since it will biodegrade under the proper conditions or can be
recycled into feedstock for packaging or composite core material; thus diverting waste
from landfills. Mycocomposites are inherently safe – the material is stable when exposed
to high temperatures (200-400C ignition point range) or ultraviolet radiation, and the use
of vegetative tissue prevents the formation of spores, a potential allergen and friable
particle if made inactive (i.e. killed) by heating within a certain time period. The
bio-based core and reinforcement can offer enhanced structural and thermal performance,
while the mycelium acts as a binder, shown in Figure 2, similar to the role of a polymer
resin matrix. A typical cycle time is about 5-7 days, as previously mentioned, to
adequately colonize the reinforcement fibers and/or core material, and longer
colonization leads to stiffer and stronger final products. A smoother surface can also be
achieved through full colonization.

(a) (b)
Figure 1: (a) Schematic and (b) picture of mycelium-based composites or
„mycocomposites‟

3
 Figure 2: SEM image of mycelium hyphae growing on a natural fiber

In parallel with the material development effort at Ecovative, the Rensselaer CATS
has been focusing on full-scale production of long fiber mycocomposites. Small batch
production of such material has already been done commercially by hand lay-up at
Ecovative [9], while the CATS has begun investigating various manufacturing processes,
particularly for the laminate skins, for manufacturing scale-up. This paper discusses
Rensselaer‟s current research related to this scale-up effort for sandwich composites.

2. CONVENTIONAL BIOCOMPOSITE MANUFACTURING

For short fiber biocomposites, conventional processes are frequently used to make
components. Bio-fibers in chopped form and plastic (petroleum- or bio-based) in the
granule forms can be extruded uniformly into pellets or logs and used in compression,
reaction injection, or injection molding.
Likewise, conventional processes are used for long fiber biocomposites. When the
thermoplastic matrix comes as a powder or solution, it is introduced into the fiber
reinforcement, fixed by heating or drying, thus making pre-impregnated materials for the
manufacture of composites by pressing or hand layup into a mold. Resin Transfer
Molding (RTM) is used to fabricate biocomposites from dry bio-fibers and natural or
synthetic resins [6]. To date, hand layup is the most widely used manufacturing method,
and products are made in small batches due to the limitations in the shape forming and

4
curing processes. This has made the cost of producing biocomposites parts remain high,
and thus limiting its extensive use in vast applications.

3. PROCESSES FOR CREATING PRE-FORMED SKINS

Similar to traditional advanced composites, mycocomposites with a sandwich
construction consist of two main parts: a single or multilayered laminate skin (woven
textiles) and loose core material mixed with mycelium. The manufacturing approach
being investigated is to create rigid skin preforms that can be easily filled with previously
colonized material that will regrow together along with growing into and integrating the
laminate skins. The general steps of this process would be: (1) cut desired skin plies in
the appropriate two-dimensional shape; (2) impregnate the skins with a natural glue; (3)
stack the impregnated skins layer by layer in the correct orientation to create the flat
laminate; (4) form the skins using molds; (5) heat and dry the formed laminate skin to
hold its shape by activating the glue and sterilize it prior to filling with loose, colonized
core material. Individual steps are discussed in more detail in the following sections.

3.1 Pre-cutting Skin Plies

There are many methods for cutting woven skin plies. The methods that were tested
with woven jute textile include hand cutting, die cutting, CNC cutting, and laser cutting.
As part of an experimental comparative study of cutting processes, a shoe-shaped ply
(shown in Figure 3(a)) was made by hand cutting, CNC cutting, and laser cutting. A
square shape was used for die cutting due to availability of a steel rule die. Process
variables monitored during the experimental investigation include capital cost (e.g.,
equipment, tooling), preparation time (time required for preparation before cutting),
cutting time, and energy consumption (estimate based on 50% of maximum power input
to machine). After each experiment, cut quality of the specimens was assessed by

5
recording the number of uncut threads, numbers of frayed threads, and cutting accuracy
on a 1-10 relative scale.

(a) (b)

(c)

(d)

Figure 3: (a) Shoe-shaped woven jute composite ply, (b) curtain-coating of an individual
ply with natural glue, (c) forming of 2-ply stack with matched molds, and (d) drying the
skins using simple heated platens.

6
3.2 Ply Impregnation with Natural Glue
Several methods were investigated for impregnating individual plies (woven jute)
with a natural glue that is compatible with mycelium growth including dip-and-soak,
spray coating, and curtain coating (shown in Figure 3(b)). For the-dip and-soak process,
two different ways were tested: dry and wet. The dry method involves putting the pre-cut
sample into the glue directly, while the wet method involves pre-wetting the sample with
water and then continuing with the dip-and-soak. For the spray coating process, a
commercially available spray gun designed for automotive painting was used, and two
different spray flow rates (corresponding to 20 psi and 30 psi input pressures) were tested.
For curtain coating, a converted fish tank waterfall filter was used. The authors found that
the contact time between the glue „waterfall‟ and the sample can be very short compared
with other impregnating methods, even though it had to be sent through twice to cover
both sides. Process outputs measured include dip time, spray pressure, sample starting
weight and wet weight, coating time, and impregnation coverage.
The natural glue used in all tests consisted of water, starch and maltodextrin. Starch
can be heat activated to create strong bonds for holding together pre-forms, but the
problem of syneresis requires the solution to be constantly mixed. Maltodextrin helps to
increase viscosity.

3.3 Forming Laminate Skins

The methods investigated for forming laminate skins into their final shape prior to
drying and sterilizing include hand layup, matched mold forming (shown in Figure 3(c)),
and double diaphragm forming (DDF), as described below:
 Hand layup requires a single mold (shoe shape in this case), although it is inherently
slower than the other two methods.
 Matched mold forming involves two matching tools (e.g., male and female) that
when fully engaged, form a gap to accommodate the thickness of the formed

7
 laminate. A wooden male mold and thermoformed plastic female mold in the shape
of a shoe sole were used to pre-form two-ply thick specimens.
 The DDF process was investigated using an existing machine available in the CATS.
DDF typically involves: sandwiching an uncured stack of reinforcement (e.g., cloth)
pre-impregnated with resin (called „pre-preg‟) between two silicone diaphragms
evacuated in between and suspending over a rigid frame; heating the pre-preg stack
(typically with a blanket heater) to lower viscosity resin; and pulling the stack down
over a rigid, stationary mold using differential pressure. Vacuum pressure on the
diaphragms helps prevent buckling of the laminate during formation. After the
vacuum between the diaphragms is released and the top diaphragm is lifted, the
pre-formed skin or core is quickly frozen or dried (mycelium can survive either
condition) to retain its formed shape and then be placed in another mold for
drying/sterilization.
Process outputs measured for these forming processes include cycle time (time required
for the forming process), glue contact time (time that a worker will be physically in
contact with the glue), degree of wrinkling, and quality on a 1-10 relative scale.

3.4 Skin Drying/Sterilization

Methods investigating for drying and sterilizing the formed laminate skins include
microwave heating, IR (infrared) oven heating, heated tooling (conduction, shown in
Figure 3(d)), IR lamp heating, a combination of a heated tool & IR lamp heating, and,
finally, air drying. Samples were not considered dry until a measurement of 0.0% content
of water was indicated by the EXTECH® MO210 moisture meter. The exception was air
dried samples which could only reach 0.6% humidity due to the lab environmental
conditions. Process outputs measured include glue usage (difference between the dry
sample weight and its weight after glue soaking), pre-preg sample weight gain (difference
between the dried sample and its original dry weight), drying time, the time used for the

8
entire drying process, and energy usage measured using a P3 international P4400
Kill-A-Watt electricity usage monitor. Compliance, strength, and heating quality were
also recorded for the drying quality.

4. COMPARISON OF MANUFACTURING METHODS

Each method was evaluated experimentally, process outputs were measured, and
methods were ranked based on a weighted scale of process variables chosen by the
authors. Ranking matrices for each of the processes described in Sections 3.1-3.4 are
provided in Tables 1-4, respectively. Note that in each of these tables, the weighted total
is calculated as the sum of each value in a column multiplied by the corresponding
weightings and then divided by the largest value in that column.
The ranking matrix for pre-cutting skin plies is shown in Table 1. In this table,
measures of cutting process quality are uncut threads and frayed threads, and data of
samples of the same cutting method are averaged. The parameter for accuracy is defined
as the relative percentage that the four sides of a cut sample match the inner sides of a
right-angle gauge. However, inaccuracy, i.e. (1 – accuracy), is used for comparison to
remain consistent with the other „smaller is preferable‟ metrics used in this table. Based
on the weighted total, die cutting method is clearly preferable to the other three methods.

Table 1: Ranking matrix for pre-cutting skin plies

Ave. Ave.
Cutting Capital Prep Ave. Inacc
Cut Time Uncut Frayed
Process Cost Time Energy uracy Weigh
Threads Threads
ted
Units $ seconds seconds Joule # # %
Total
Weighting
*8/25000 *6/125 *10/177 *5/321600 *8/20 *5/10 *4/60
(1-10)
Hand Cut $10.00 125 177 0 1.9 9 40 23.93
Die Cut $5,000.00 5.58 4.066 3654 0.5 0.2 0 2.60
Laser Cut $10,000.00 60 64 59,520 20 10 0 29.12
CNC Cut $25,000.00 60 48 321,600 10 5 60 31.84

9
 The ranking matrix for ply impregnation with natural glue is shown by Table 2. It
should be mentioned that the cost of glue is the average unit price multiplied by the
average glue usage. Based on the weighted totals, the dip & soak and curtain coating
methods are preferable to all other impregnation methods.

Table 2: Ranking matrix for individual ply impregnation

Impregnation
Ave. Cycle Time Ave. Uncovered Cost of Glue
Processes
Weighted Total
Units seconds % $
Weighting (1-10) *10/56 *8/100 *5/0.0905
Dip & Soak dry 10.83 0.00 $0.0905 6.93
Dip & Soak wet 15.00 0.00 $0.0754 6.84
Spray Coating
52.50 12.50 $0.0658 14.01
(30psi)
Spray Coating
56.00 35.00 $0.0592 16.07
(20psi)
Curtain Coating
7.17 3.33 $0.0671 5.25
(twice)

The ranking matrix for forming laminate skins is shown in Table 3. The DDF process
was immediately eliminated from consideration, because it simply was not suited for the
relatively messy „glued‟ materials or for the low cycle times envisioned for production.
With regards to ranking metrics, “Inferiority” is used instead of quality, since the ranking
required „smaller is preferable‟ metrics. Based on the ranking, match mold forming is the
clear choice for high production.
Table 3: Ranking matrix for forming laminate skins
Stacking Degree of Inferiority
Cycle time Contact time
Processes Wrinkling (1/quality)
Weighted
Units min seconds 0-10 >0.1
Total
Weighting
*10/8 *8/65 *8/5 *8/0.167
(1-10)
Hand layup 8 65 5 0.167 34
Match mold
1 20 1 0.111 10.63
stamping

10
 The ranking matrix for skin drying is shown in Table 4. The strength of the samples
is converted to peel weakness for statistical purposes. The strength of the sample is
defined as the maximum force required to separate two 2.5 cm-wide plies glued together.
And the peel weakness is the reciprocal, namely 2.5 cm width divided by the maximum
forces to separate two plies. The compliance of the samples is the displacement of a 2.5
cm-wide cut sample in three-point bending experiment with an 8-cm span length and
subjected to a 292 g weight. The overall drying quality is measured on a relative scale
between -10 and 10, where a negative number means under-heated and white (signifying
starch did not fully activate), a positive number stands for overheated and burnt, and 0
signifies an optimal glued condition. Absolute values are used in the comparison, since
negative numbers would not work for the weighted process output summation. After
comparison, the authors concluded that low powered microwave drying and flipped
sample method are of relatively low total number and were selected for further
investigation.

Table 4: Ranking matrix for skin drying

Ave. Ave. Overall Overall
Drying Ave. Cycle Energy Overall
Glue Weight Peel Drying
Methods Time Usage Compliance
Usage Gain Weakness Quality
(2.5cm (-10~10,
(displaceme width where -10
nts in 8cm divided by is Weigh

minutes span 3-point maximum underheate ted

Units g/ply g/ply J/ply
per ply bending with force to d white and Total
a 292.34g separate 2 10 is
weight) mm plies) overheated
cm/lb. burnt)

Weighting *3/12.9 *5/4.837

*10/720 *6/485000 *7/10 *7/9.259 *9/|3|
(1-10) 99 5
Microwave
11.704 3.822 2.5 8.06E+04 6 0.725 |0.75| 14.7
P50
Microwave
12.407 4.727 3.3 8.60E+04 2 0.315 |0.25| 11.3
P40

11
Convection
12.999 4.608 5.1 2.78E+05 2.5 0.511 |1| 16.4
Oven
Heated Tool 11.7625 3.9775 10 4.85E+05 3 1.969 |0| 16.6
Infrared
11.645 4.07 20 2.93E+05 10 7.353 |-1| 26.4
light
Heated tool
12.014 3.881 5.892 3.72E+05 2 0.564 |-2.5| 20.8
& Infrared
Heated tool
& Infrared 12.459 4.442 6.65 4.20E+05 2 0.912 |0.25| 15.6
(flipped)
~0.5
Air dry 11.2325 4.8375 0.00E+00 7 9.259 |-3| 38.5
day=720

5. CONCLUSIONS AND FUTURE WORK

This paper describes the evaluation of several manufacturing processes being
considered for large-scale production of a novel mycocomposite material that is grown
instead of synthesized using conventional processes. Based on the experimentally based
evaluations shown in Tables 1-4, a more refined manufacturing approach is suggested for
further consideration. A flowchart of the overall approach is shown in Figure 4. With a
desired cutting die, woven cloth is cut into the exact shape that will later fit the mold. The
pre-cut skin plies then feeds through a glue impregnation station involving either
dip-and-soak coating or curtain coating. Glue-impregnated skin plies will then be stacked
manually and formed to a desired shape by matched tool stamping. The shape added wet
skin plies will then be dried either by low-power microwave or heated tooling augmented
with IR lamps. Once the pre-formed and dried/sterilized skins are completed, they are
filled with inoculated core material and allowed to grow into a finished mycomposite
sandwich structure. Future work will involve prototyping the entire manufacturing line
and testing it with a realistic part shape to determine the best combination of processes
and automation for high-volume production.

12
 Figure 4: Proposed layout of manufacturing system concept

REFERENCES
1. Fowler, Paul A, J Mark Hughes, and Robert M Elias. "Biocomposites from crop
fibres and resins." IGER INNOVATIONS. 2007: 66-68. Web. 13 Jan. 2013.
<http://www.aber.ac.uk/en/media/fowlerhugheselias.pdf>.
2. Mohanty, A. K., M. Misra, and L.T. Drzal. "Sustainable Bio-Composites from
Renewable Resources: Opportunities and Challenges in the Green Materials
World." Journal of Polymers and the Environment. 10.1/2 (2002): 19-26. Web. 13
Jan. 2013.
3. SAHEB, D. NABI, and J.P. JOG. "Natural Fiber Polymer Composites: A
Review." Advances in Polymer Technology. 18.4 (1999): 351-63. Web. 13 Jan.
2013.
4. William F. Powers, “Automotive Materials in the 21st Century”, Adv. Mater.
Process, May 2000, p. 38-41.
5. G. H. Brother and L. L. McKinney (1940) Ind. Eng. Chem. 32, 1002.
6. Drzal, Lawrence T., A.K. Mohanty, and M. Misra. "BIO-COMPOSITE

13
 MATERIALS AS ALTERNATIVES TO PETROLEUM-BASED COMPOSITES
FOR AUTOMOTIVE APPLICATIONS." 2001 SPE ACCE PROCEEDINGS Web.
13 Jan. 2013. <http://speautomotive.com/SPEA_CD/SPEA2001/pdf/e/E1.pdf>.
7. Flanigan, Cynthia, Christine Perry, Ellen Lee, Dan Houston, Debbie Mielewski,
and Angela Harris. “Use of agriculture materials in flexible polyurethanes and
composites for automotive applications.” 2006 SPE ACCE PROCEEDINGS Web.
13 Jan. 2013.
<http://speautomotive.com/SPEA_CD/SPEA2006/PDF/c/c3.pdf>.
8. "Our Technology." Ecovative. N.p.. Web. 25 Mar. 2013.
<http://www.ecovativedesign.com/about-our-materials/core-tech/>.
9. "About Ecovative." Ecovative. N.p.. Web. 13 Jan. 2013.
<http://www.ecovativedesign.com/about-ecovative/>.

14

View publication stats