**4. Results**

## *4.1. Fabrication*

We successfully developed a custom robotic fabrication process consisting of two steps for this study: robotic wood fiber laying and ultrasonic wood welding. These individual processes were carried out, respectively, using different end effectors mounted on a robot arm. By carefully placing the wood fibers and binding them where necessary, we produced flat lattices made up of two orthogonal layers with this method and used them to reinforce mycelium blocks. Thanks to the mechanical properties and directionality of the material being preserved with this fabrication method, precise control over the reinforcement orientation was achieved.

Despite the geometric freedom of this additive manufacturing method, we encountered some challenges in the production: when placing the veneer, the start and end points of the strips should be fixed. Similarly, when more than one layer is deposited on one point, the strips must be bound together to stay in place until mycelium growth. Since only 2D lattices were produced for this study, this was solved by fixing the end points of the veneers with double sided tape on an aluminum plate. However, to produce more complex geometries with multiple layers, a more robust and local solution needs to be researched.

As a result of the initial welding tests, the most promising bond was obtained with the sample where the wood side of one veneer was welded to the fleece side of the other. Therefore, all samples were produced, while keeping the wood side on top; two layers of veneer were overlapped on their opposite sides (wood to fleece) and prevented the fleece from sticking to the sonotrode.

The welds on the intersection points were successful in keeping the lattice together during mycelium growth. However, not every point could be welded at once, or with the same quality due to slight pressure differences caused by using robot motion to apply pressure for welding. The welding setting used for most of the intersection points was 0.5 seconds with 100% amplitude. While this setting performed well in most welds, inconsistencies were observed, with burned welds and welding failures. In our robot setup, we used contact pressure through robot motion, which was not possible to be precisely controlled. This is therefore assumed to be one of the main reasons for welding inconsistencies. With the welding equipment being highly sensitive, if the sensors detect that the target welding time is not reached, or the system uses too much power to weld, the process is interrupted. In further studies, the integration of a force-controlled pneumatic cylinder and motion control system is planned to ensure constant pressure, prevent delays in production, and provide weld consistency. pressure for welding. The welding setting used for most of the intersection points was 0.5 seconds with 100% amplitude. While this setting performed well in most welds, inconsist‐ encies were observed, with burned welds and welding failures. In our robot setup, we used contact pressure through robot motion, which was not possible to be precisely con‐ trolled. This is therefore assumed to be one of the main reasons for welding inconsisten‐ cies. With the welding equipment being highly sensitive, if the sensors detect that the tar‐ get welding time is not reached, or the system uses too much power to weld, the process is interrupted. In further studies, the integration of a force‐controlled pneumatic cylinder and motion control system is planned to ensure constant pressure, prevent delays in pro‐ duction, and provide weld consistency.

### *4.2. Testing*

### 4.2.1. Tensile Tests *4.2. Testing* 4.2.1. Tensile Tests

Most samples showed wood tensile failure (Figure 6b), rather than welding area failure, proving that the weld strength is higher than the veneer tensile strength. The average tensile strength of the welded samples was measured as 63.6 ± 10.5 MPa, which is almost the same as the average tensile strength of a single maple veneer strip, 61.95 ± 10.83 MPa, and confirms the previous statement (Figure 6a). Most samples showed wood tensile failure (Figure 6b), rather than welding area fail‐ ure, proving that the weld strength is higher than the veneer tensile strength. The average tensile strength of the welded samples was measured as 63.6 ± 10.5 MPa, which is almost the same as the average tensile strength of a single maple veneer strip, 61.95 ± 10.83 MPa, and confirms the previous statement (Figure 6a).

**Figure 6.** Tensile strength tests: (**a**) Comparison graph of maple veneer's tensile strength to welded joints' tensile strength; (**b**) Tested welded veneer samples. **Figure 6.** Tensile strength tests: (**a**) Comparison graph of maple veneer's tensile strength to welded joints' tensile strength; (**b**) Tested welded veneer samples.

### 4.2.2. Compression Tests 4.2.2. Compression Tests

The results of compression, pull‐out and flexural tests of all samples, together with their densities, are summarized in Table 1. The results of compression, pull-out and flexural tests of all samples, together with their densities, are summarized in Table 1.

 The average compressive strength of the samples was 1.2 MPa and the average elastic modulus in compression was measured to be around 4.1 MPa. As it can be seen in Figure 7a, the samples were compressed until reaching 50% of their original height. No cracking or breakage was observed during the tests. The samples showed deformation under compression load and were compressed until the end of the test. It was also observed, when the test continued beyond 50% of the sample's original height, that higher loads could be achieved. However, given the recommendations of the ASTM D3574, the test continued until the samples were compressed to 50% of their original height and showed similar behavior to conventional foams and flexible cellular materials.


**Table 1.** Summary table for physical and mechanical properties of mycelium-based composites with and without veneer lattices.

\* Flexural failure was not observed for these samples; the failure mode was shear as explained in the text.

**Figure 7**. Samples during testing and after failure: (**a**) Compressive strength test; (**b**) One-side veneer pull-out test; (**c**) Two-side unwelded veneer pull-out test; (**d**) Two-side welded veneer tensile test. **Figure 7.** Samples during testing and after failure: (**a**) Compressive strength test; (**b**) One-side veneer pull-out test; (**c**) Two-side unwelded veneer pull-out test; (**d**) Two-side welded veneer tensile test.

However, for samples prepared with welded overlapped maple veneers, the failure of the veneer strips due to tensile mode of failure was observed, before the veneer could be pulled out (Figure 7d). This can be explained by the higher failure load observed during the tests compared to the other pull-out samples. Furthermore, when the resulting stress is compared with the tensile strength of the maple veneer, it can further validate the hypothesis that the weld strength is relatively higher than the bond strength between the

### 4.2.3. Pull-Out Tests

The pull-out samples with one strip of maple veneer in the middle showed similar IFSS as those samples prepared with two strips of unwelded maple veneer. Mycelium growth was observed on the maple veneer surface that was embedded within the mycelium matrix (Figure 7b,c). The samples have shown a relatively good bonding: the bond strength between the mycelium matrix and the veneer strips was measured as 0.34 MPa and 0.36 MPa for a single veneer strip and overlapped veneer strips, respectively. The slight increase in the IFSS in samples with two unwelded overlapped veneer strips can be attributed to the better growth of the mycelium network around and between the layers of the veneer strips. A stronger bond was developed in these areas in comparison to the samples with a single strip of veneer where less surface area resulted in a lower mycelium growth density. For both series of samples, clear pull-out of veneers from the mycelium matrix was observed.

However, for samples prepared with welded overlapped maple veneers, the failure of the veneer strips due to tensile mode of failure was observed, before the veneer could be pulled out (Figure 7d). This can be explained by the higher failure load observed during the tests compared to the other pull-out samples. Furthermore, when the resulting stress is compared with the tensile strength of the maple veneer, it can further validate the hypothesis that the weld strength is relatively higher than the bond strength between the veneer strips and the mycelium matrix (Figure 6). However, further testing is required to find out the effect of mycelium growth combined with welded veneers on improving the bond strength in wood-veneer-reinforced mycelium-based composites.

## 4.2.4. Flexural Tests

Lightweight samples with one layer of high-density veneer lattice and the ones with top and bottom low-density veneer lattices showed shear mode of failure, while all the other samples, including dense boards with low- and high-density veneer lattices, and samples with no veneer lattices showed flexural mode of failure (Figure 8). The flexural strength of lightweight blocks increased slightly with the addition of one layer of low-density veneer lattice in the middle, compared to non-reinforced blocks. On the other hand, the use of high-density lattices resulted in shear failure and lower flexural strength. Similarly, samples with top and bottom low-density lattices also showed lower flexural strength. The results are summarized Figure 9.

The abovementioned behavior could be explained by the shear failure mode as a result of the potentially lower bonding strength between the veneer lattices and mycelium matrix. High-density lattices decrease the areas where the mycelium network would grow through the lattice holes and connect the two sides of the block divided by the lattice. This would create a weaker interlocking mechanism, which could result in a higher chance of de-bonding when exposed to flexural loads. Furthermore, the lower flexural strength of these samples compared to low-density lattices could also be attributed to the mycelium growth on the veneer lattices. In the case of high-density lattices, more surface area would result in a higher bonding strength between the veneer strips and substrates by forming a stronger mycelium network. However, as was observed from the pull-out tests of samples with unwelded overlapped veneer strips, the bond mechanism was not fully developed and all the samples showed clear pull-out failure, rather than the tensile failure of the veneer strips. Therefore, it is possible to state that weaker bonding areas between the mycelium matrix and the veneer strips increase the chance of de-bonding and interlaminar shear failure.

find out the effect of mycelium growth combined with welded veneers on improving the

Lightweight samples with one layer of high-density veneer lattice and the ones with top and bottom low-density veneer lattices showed shear mode of failure, while all the other samples, including dense boards with low- and high-density veneer lattices, and samples with no veneer lattices showed flexural mode of failure (Figure 8). The flexural strength of lightweight blocks increased slightly with the addition of one layer of lowdensity veneer lattice in the middle, compared to non-reinforced blocks. On the other hand, the use of high-density lattices resulted in shear failure and lower flexural strength. Similarly, samples with top and bottom low-density lattices also showed lower flexural

bond strength in wood-veneer-reinforced mycelium-based composites.

4.2.4. Flexural Tests

strength. The results are summarized Figure 9.

**Figure 8**. Samples during testing and after failure: (**a**) lightweight block under 3-point flexural test; (**b**) Block without or with low-density lattice in the middle; (**c**) Block with high-density lattice in the middle (shear failure); (**d**) Block with two layers of low-density lattices close to the top and bottom of the block; (**e**) Dense board under 3-point flexural test; (**f**) Dense board after failure without lattice reinforcement. **Figure 8.** Samples during testing and after failure: (**a**) lightweight block under 3-point flexural test; (**b**) Block without or with low-density lattice in the middle; (**c**) Block with high-density lattice in the middle (shear failure); (**d**) Block with two layers of low-density lattices close to the top and bottom of the block; (**e**) Dense board under 3-point flexural test; (**f**) Dense board after failure without lattice reinforcement. *Biomimetics* **2022**, *7*, x FOR PEER REVIEW 16 of 21

**Figure 9**. Flexural properties, including strength and elastic modulus: (**a**) Flexural strength of light‐ **Figure 9.** Flexural properties, including strength and elastic modulus: (**a**) Flexural strength of lightweight blocks; (**b**) Elastic modulus in flexure of lightweight blocks; (**c**) Flexural strength of dense boards; (**d**) Elastic modulus in flexure of dense boards. \* Shear failure was observed, further explanation is given in the text.

weight blocks; (**b**) Elastic modulus in flexure of lightweight blocks; (**c**) Flexural strength of dense boards; (**d**) Elastic modulus in flexure of dense boards. \* Shear failure was observed, further expla‐

sult of the potentially lower bonding strength between the veneer lattices and mycelium matrix. High‐density lattices decrease the areas where the mycelium network would grow through the lattice holes and connect the two sides of the block divided by the lattice. This would create a weaker interlocking mechanism, which could result in a higher chance of de‐bonding when exposed to flexural loads. Furthermore, the lower flexural strength of these samples compared to low‐density lattices could also be attributed to the mycelium growth on the veneer lattices. In the case of high‐density lattices, more surface area would result in a higher bonding strength between the veneer strips and substrates by forming a stronger mycelium network. However, as was observed from the pull‐out tests of sam‐ ples with unwelded overlapped veneer strips, the bond mechanism was not fully devel‐ oped and all the samples showed clear pull‐out failure, rather than the tensile failure of the veneer strips. Therefore, it is possible to state that weaker bonding areas between the mycelium matrix and the veneer strips increase the chance of de‐bonding and interlami‐

Unlike the lightweight samples, dense boards showed clear flexural failure, which indicates a stronger bond mechanism between the veneer strips and mycelium matrix (Figure 8f). Even though no shear failure was observed, similar trends in flexural strength could be observed when the high‐density veneer lattices were used in dense boards. No

nation is given in the text.

nar shear failure.

Unlike the lightweight samples, dense boards showed clear flexural failure, which indicates a stronger bond mechanism between the veneer strips and mycelium matrix (Figure 8f). Even though no shear failure was observed, similar trends in flexural strength could be observed when the high-density veneer lattices were used in dense boards. No significant increase in flexural strength was detected within dense boards with high-density lattices compared to dense boards with no lattices. However, dense boards with low-density lattices showed a significant increase in both flexural strength and elastic modulus. In general, it was also observed that the increase in density helped to increase the flexural properties when the results of dense boards are compared with lightweight blocks.

Further testing is required to evaluate the impact of different veneer lattice densities and layouts in combination with different substrate densities on the flexural properties of wood-veneer-reinforced mycelium-based composites. Furthermore, the investigation of veneer placement within the samples along the height of the blocks should also be carried out to explore the bond mechanism developed between mycelium matrix and the veneer lattices with varying densities, and their impacts on flexural properties of the final samples.

Lightweight samples reinforced with either one layer of high-density lattice in the middle or two layers of low-density lattices on top and bottom of the block showed shear failure (Figure 8c,d). However, for lightweight samples reinforced with one layer of lowdensity veneer lattice and dense boards with low- and high-density veneer lattices, flexural failure was observed as the dominant mode of failure as expected (Figure 8b,f).

### **5. Discussion**

The wood fiber laying process used in this study has the potential to become a resourceefficient, rapid production method, with material carefully placed in the structurally required areas. Fiber placement and binding were carried out with two different end effectors, as sequential steps of the process. In order to speed up production, further studies are planned to develop a single tool that can lay wood veneers and at the same time weld the intersections.

The novel wood-veneer-reinforced mycelium composites developed in this study were investigated for their mechanical properties, including compressive strength, pull-out strength and flexural properties. The suitability of 2D veneer lattices as a reinforcement system with welded and unwelded joints was also investigated separately through a series of tensile tests. The results of the investigation of welded joints show that they perform relatively well; in the majority of the tensile tests, no failure in the joints was observed, which indicates that the joints have a higher strength than the veneer itself. The bonding between the veneer strip and mycelium matrix was investigated through a series of pull-out tests and the results show that the bond might not have been developed fully, as most of the single veneer-strip-reinforced cubes showed purely pull-out failure modes rather than any failure in the veneer strip. However, the visual examination after completion of the tests confirmed the growth of mycelium network on the veneer strips. This again validates our hypothesis that the selected mycelium species (*Ganoderma lucidum*) can grow well when combined with the selected veneer species (maple) and hemp hurds as the main substrate.

Furthermore, the performance of the welded and unwelded joints was investigated through a series of pull-out tests with similar overlapping veneer areas embedded within the mycelium matrix. It was observed that the welded joints outperform the unwelded ones. The results show that, even though the mycelium growth was observed in both cases on the veneers, the interfacial shear strength developed within the unwelded veneer strips and mycelium matrix was lower than the strength of the welded joints. Further investigation on enhancing the mycelium growth on the veneer strip and improving the interfacial shear strength between the mycelium matrix and veneer strip is necessary to achieve a better bonding strength.

The results of the flexural tests on various samples once again strengthen the hypothesis that the bonding between the veneer reinforcement and mycelium matrix plays an important role in the structural integrity and mechanical properties of these composites.

Moreover, compressing the lightweight blocks into dense boards showed a significant improvement in flexural properties as a result of densification, and improved the bending mechanism between the veneer lattice and mycelium matrix. While samples with top and bottom veneer reinforcement did not show any significant increase in the overall flexural properties, samples with one layer of low-density veneer lattice before and after compression showed better flexural performance. The lower flexural strength and elastic modulus measured correspond to the shear mode of failure observed during the tests. Therefore, further investigation is necessary to identify the optimum design of the veneer lattices and to explore the effect of connecting the top and the bottom reinforcement lattices, namely 3D lattices. Moreover, compressing the lightweight blocks into dense boards showed a significant im‐ provement in flexural properties as a result of densification, and improved the bending mechanism between the veneer lattice and mycelium matrix. While samples with top and bottom veneer reinforcement did not show any significant increase in the overall flexural properties, samples with one layer of low‐density veneer lattice before and after compres‐ sion showed better flexural performance. The lower flexural strength and elastic modulus measured correspond to the shear mode of failure observed during the tests. Therefore, further investigation is necessary to identify the optimum design of the veneer lattices and to explore the effect of connecting the top and the bottom reinforcement lattices, namely

*Biomimetics* **2022**, *7*, x FOR PEER REVIEW 18 of 21

Composed of one bottom and one top 2D lattice connected by an undulating layer of wood veneer, 3D lattice reinforcements could potentially improve the shear capacity of the mycelium composites and provide additional strength and stiffness via the spatial lattice system. Their design would be strictly connected to the design of the base 2D lattice. Following the production of the flat lattice, the layer that gives the structure its depth would be achieved through placing the material diagonally between the two opposite corners of a quadrilateral cell created by the 2D lattice, with a pre-calculated length. More material length would result in more structural depth. 3D lattices. Composed of one bottom and one top 2D lattice connected by an undulating layer of wood veneer, 3D lattice reinforcements could potentially improve the shear capacity of the mycelium composites and provide additional strength and stiffness via the spatial lat‐ tice system. Their design would be strictly connected to the design of the base 2D lattice. Following the production of the flat lattice, the layer that gives the structure its depth would be achieved through placing the material diagonally between the two opposite cor‐ ners of a quadrilateral cell created by the 2D lattice, with a pre‐calculated length. More material length would result in more structural depth.

This system could also achieve surfaces with varying depths through the management of the middle layer's height and the corresponding non-parallel and non-planar top and bottom surfaces. Similarly, through the gradual cell size modifications along the structure, heterogeneous reinforcement could be achieved. The parameters that can be adjusted on a cell level provide high flexibility, and opportunities for lightweight, optimized reinforcement based on local requirements. This system could also achieve surfaces with varying depths through the manage‐ ment of the middle layer's height and the corresponding non‐parallel and non‐planar top and bottom surfaces. Similarly, through the gradual cell size modifications along the structure, heterogeneous reinforcement could be achieved. The parameters that can be adjusted on a cell level provide high flexibility, and opportunities for lightweight, opti‐ mized reinforcement based on local requirements.

However, certain limitations must be considered when designing 3D lattices with wood veneers: different veneer thicknesses and species allow for different bending radii. Due to the fibers being oriented in the direction of fiber laying, forcing the material into a very small bending radius would result in the filament breaking. Therefore, the minimum bending radius would be the main determining factor for the component height and as a result, the 3D lattice density. In Figure 10, some 3D lattice design studies based on the bending radius of maple veneer can be seen. However, certain limitations must be considered when designing 3D lattices with wood veneers: different veneer thicknesses and species allow for different bending radii. Due to the fibers being oriented in the direction of fiber laying, forcing the material into a very small bending radius would result in the filament breaking. Therefore, the minimum bending radius would be the main determining factor for the component height and as a result, the 3D lattice density. In Figure 10, some 3D lattice design studies based on the bending radius of maple veneer can be seen.

**Figure 10.** 3D lattice layout studies (dimensions in mm). **Figure 10.** 3D lattice layout studies (dimensions in mm).

**6. Conclusions** 

### **6. Conclusions**

A novel wood-veneer-reinforced mycelium-based composite material was developed for this study as a sustainable and green alternative to traditional building materials with potential applications in the construction industry. Structural testing on physical prototypes was carried out to investigate the fundamental mechanical properties of this novel composite. The test samples were prepared with different variations of veneer lattices as reinforcement systems and tested for compressive strength, bond strength, and flexural properties. The tests provided an initial understanding of the mechanical behavior of the wood-veneer-reinforced mycelium composites in terms of densities, strength and stiffness at material scale. Both strategies, integrating a topologically designed veneer lattice and compression with heat and pressure, proved to be effective methods in increasing the bending resistance of the presented composites. It was shown that the effect of veneer lattices as reinforcement systems is strictly tied to the density and configuration of the lattice. For lightweight blocks, the most promising results were achieved with a single layer of low-density veneer lattice placed in the middle of the mycelium block. This configuration helped to increase the flexural strength of the block slightly (approximately from 0.17 MPa to 0.19 MPa), whereas the high-density lattice and two low-density lattices at the top and bottom of the block resulted in a lower flexural strength (approximately 0.16 MPa and 0.13 MPa, respectively) than that of the unreinforced block itself. The samples with two low-density lattices demonstrated a low flexural strength and elastic modulus and resulted in shear failure. Therefore, 3D lattice systems connecting top and bottom lattices are proposed to avoid shear failure in lightweight blocks for future studies.

The dense boards with one low-density lattice in the middle demonstrated a similar trend to the lightweight blocks and increased the flexural strength to more than double (from approximately 10 MPa to 25 MPa) of the unreinforced dense boards. On the other hand, the dense boards with one high-density lattice in the middle did not show a significant change in flexural strength compared to the unreinforced dense boards.

When the two methods are compared, dense boards have a better overall flexural strength. The dense boards reinforced with one low-density lattice are the most promising specimens, and would be appropriate for applications that require planar components and higher bending resistance. However, if more complex geometries that do not require high bending resistance are needed, lightweight blocks reinforced with one low-density lattice would be suitable.

The study provided the fundamental material inputs for the further development of the system at a larger scale. In the next steps, a digital model will be developed to integrate the material properties as design inputs and material constraints; geometrical variations as design variables; and structural Finite Element Analysis (FEA) and acoustic analyses as solvers to evaluate and optimize various design options within one digital computational framework.

Further studies, including the investigation of the growth compatibility between other wood veneers and mycelium species combined with a range of available organic waste by-products from wood and agricultural industries, will be carried out in the next steps of the research. Additional mechanical testing of the mycelium composites as larger panels reinforced with 3D lattice systems made of veneer are also planned to gain further insights into the materials' behavior, which will subsequently support the design and development of these composites. The initial results obtained show that wood-veneer-reinforced mycelium composites could be a promising environmentally friendly and sustainable substitute material to conventional building materials with potential applications in the context of architecture.

**Author Contributions:** Conceptualization, E.Ö., A.J. and N.S.; methodology, E.Ö., A.J. and A.R.; software, A.R. and E.Ö.; validation, N.S., A.R. and E.Ö.; formal analysis, A.J. and N.S.; investigation, A.J.; data curation, N.S.; writing—original draft preparation, E.Ö. and A.J.; writing—review and editing, N.S., A.R., N.N., S.R., A.D., I.A. and P.E.; visualization, E.Ö., N.S. and N.N.; supervision, P.E., D.E.H. and J.W.; project administration, E.Ö., A.R. and P.E.; funding acquisition, P.E., D.E.H. and J.W. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by "Forschungsinitiative Zukunft Bau des Bundesinstitutes für Bau-, Stadt- und Raumforschung", grant number 10.08.18.7-21.48.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data are available upon request.

**Acknowledgments:** The authors would like to thank Jannis Heise from the Fachgebiet Trennende und Fügende Fertigungsverfahren (TFF) at the University of Kassel for conducting tensile strength tests on single maple veneer strips and welded maple veneer joints, and analyzing and visualizing the test results; Thomas Bierwirth and Marco Klocke from Heitz Furnierkantenwerk, Melle for supplying wood veneers for physical prototyping and testing; and lastly, Tobias Neff and Christian Ringwald from Weber Ultrasonics, Karlsbad for technical support and ultrasonic welding equipment supply.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

### **References**

