Progress 09/30/01 to 10/01/07
Outputs
OUTPUTS: The most significant outputs of this project were: development of a beta version user-friendly computer software to analyze the bending properties of Parallel Strand Lumber (PSL) based on strand properties; simulation and experimental validation of this software on full-size PSL beam tests; use of this software to predict the effect of void content in PSL; and development of a novel technique to characterize void distribution of PSL. The methodology developed in this project continues to be refined; work is ongoing today through a collaborative effort with the department of Civil and Environmental Engineering, UMass. The project facilitated the higher level education of two graduate students and involved four undergraduate students. The work was disseminated through three refereed journal publications and resulted in two conference papers and presentations.
PARTICIPANTS: Dr. Peggi Clouston. Principal investigator. Dr. Clouston trained, directed and supervised students in conducting the activities of the study. She wrote computer code and implemented physical characteristics of Parallel Strand Lumber into the computer model. Ms. Sufen Liu. Masters of Science student. Ms. Liu carried out all preparatory tasks, took measurements and conducted the full-size experimental tests. She performed statistical analysis on the experimental data and used the software program COMAP to carry out computer simulations. Mr. Alexander Schreyer. Research Assistant. Mr. Schreyer integrated the initial computer program (FORTRAN 77 base) into a user-friendly Windows input/output interface. Ms. Tanya Favorite. Senior undergraduate student. (Stipend) Ms. Favorite conducted experimental tests on representative volume elements (RVEs) of Southern Pine Parallel Strand Lumber (PSL). Mr. Christopher Phelps. Senior undergraduate student. (Stipend) Ms. Favorite
conducted experimental tests on representative volume elements (RVEs) of Southern Pine Parallel Strand Lumber (PSL). Mr. Ryan Mador. Senior undergraduate student. (Stipend) Mr. Mador helped to prepare and maintain equipment for conditioning of test specimens.
TARGET AUDIENCES: The target audience is the general wood science community and Wood Composite manufacturers.
Impacts
The resources provided and the activities conducted in this study has led to the contribution of a novel, new, and sorely needed, methodology and software tool to simulate the bending properties of structural wood composites. For versatility, the model is based on the properties of the elements that make up the composite. Although still in beta-testing stage, the computer model is proving to be a promising tool for analyzing and developing future wood composites. Ultimately, the software is intended to be used by inventors, manufacturers and academics to change the current practice of wood composite development. The software may be used as a quick and inexpensive design tool to refine and optimize new wood composites; for example, to enhance wood fiber sustainability or to develop less expensive wood products.
Publications
- Clouston, P. 2007. Characterization and Strength Modeling of Parallel Strand Lumber. Journal Holzforschung, Vol. 61, pp. 394-399
|
Progress 10/01/05 to 09/30/06
Outputs
This past year capitalized on the progress made from the previous year toward a user-friendly computer model - now coined COMAP (COMposite Analysis Program). COMAP was used to explore two critical issues pertaining to the project: 1) incorporation of voids in the analysis of Parallel Strand Lumber, PSL (part of objective 1) and 2) prediction of full-size PSL (objective 2). In addressing issue 1, the results from the digital image analysis, carried out in 2003/2004, were used to facilitate a numerical evaluation of three different void distributions (i.e. low void content, high void content and simulated PSL void content). The three void distributions were compared to predict the effect of reducing and increasing void content in PSL; as expected, larger void content resulted in lower capacities. Exact relationships were elucidated through cumulative probability distributions and average percent differences. The void content of PSL can be reduced by carefully altering
process parameters, such as pressure and strand alignment, but with a higher cost to manufacture. This study on void influence provided insight into the trade-off between cost and desired product performance. The results were presented at the World Conference on Timber Engineering in Portland, OR and published in the conference proceedings. Issue 2 was the thesis topic of Masters student, Sufen Liu. In this work, COMAP was used to simulate the load-displacement curve of PSL beams in full-size three-point bending. Five hundred replications were generated and compared to the results of the experimental tests conducted in 2004/2005 for model validation. Using the raw data from the representative volume element tests performed in 2003/2004, a significant size effect was found to exist. A parametric study was then conducted to quantify the size effect. Upon calibration of the raw data, excellent agreement between predicted and experimental data was found. (Grad students = 1)
Impacts
This computer model will be a long-awaited tool for inventors and manufacturers of wood composites to optimize and guide development of new products. The model will foster innovation with far reaching environmental and economic impact. With this tool, for example, manufacturers will be able to quickly and inexpensively investigate: 1)the use of weed species in wood composites and thereby reduce fuel for wildfires or 2) the use of cheaper, juvenile wood to gain greater economic advantages.
Publications
- CLOUSTON, P.; LIU, S. 2006. Predicting the Influence of Macro-Void Distribution in Parallel Wood Strand Composites. World Conference of Timber Engineeering, Portland, Oregon. (paper and presentation)
|
Progress 10/01/04 to 09/30/05
Outputs
Progress has been made on several aspects of the project this past year. 1) The computer program (FORTRAN 77 base) has been integrated into a user-friendly Windows input/output interface, partially completing objective 4. This work was completed at this stage for convenience and efficiency. This improvement enables the programmer to input data and read output data (necessary throughout the project) more efficiently. 2) The computer model was then used to predict the initiation and propagation of failure of small-scale PSL specimens under tension, compression and 3 point bending. Plots were created to study the spread of the plastic zones for varying mean applied stress to yield stress ratio, for both an isotropic case and an orthotropic case. This research was carried out to get a fundamental understanding of how the orthotropic nature of composites (in general) influences the progression of failure through a member. This type of knowledge can aid manufacturers in the
design of more reliable wood composites. 3) Full-size Southern Pine Parallel Strand Lumber (PSL) beams were tested in three-point bending. This data is necessary to complete the second objective.
Impacts
This years work made significant in-roads towards a user-friendly computer model to simulate the mechanical properties of wood based composites. The program was used to provide fundamental knowledge of wood composite behavior which can lead to more environmentally responsible wood products. The economic impact of the model is the time and cost savings it can provide to wood composite manufacturers.
Publications
- No publications reported this period
|
Progress 10/01/03 to 09/30/04
Outputs
In this past year, two experimental programs were conducted: 1) five tests, each with 50 replications, were performed on representative volume elements (RVEs) of Southern Pine Parallel Strand Lumber (PSL) to establish mechanical properties in parallel-to-grain tension, perpendicular-to-grain tension, parallel-to-grain compression, perpendicular-to-grain compression, and shear; and 2) an extensive digital image analysis was carried out to scientifically ascertain the macro-void distribution of Southern Pine PSL. The first series of experimental tests were required to initiate the second objective of the project (ie. the creation of superelements for prediction of full-size wood composite members). The resulting database of constitutive properties of RVEs will be used to verify computer simulations of small beam elements (ie. superelements) which are yet to be performed in the substructuring analysis of the computer program. The second experimental program was conducted
to enhance the data obtained in the previous year for void distribution of PSL. A more thorough approach was employed, using digital analysis software, to characterize the macro-void distribution over a surface area of almost 8 times that of the previous year's analysis. The computer program, Matlab, was used to read, enhance and process 750 digital photographs whereby void area was distinguished from solid wood area through color contrasting techniques.
Impacts
This project will produce a user-friendly computer model to simulate the mechanical properties of wood based composites. The impact of this model is both economic (it will save time and money in development of new wood composites) and environmental (it will aid in design of wood composites and thereby lead to better utilization of wood and wood fibre).
Publications
- Clouston, P. 2004. Computational Modeling of Parallel Strand Lumber. World Conference of Timber Engineeering, Lahti, Finland
- Clouston, P. 2003. A Stochastic Plasticity Approach to Strength Modeling of Wood Composites. Wood Composite Symposium, Pullman, Washington, USA
|
Progress 10/01/02 to 09/30/03
Outputs
Two defining characteristics of PSL (grain angle and void distribution) were measured and included as random variables into the computer model. Discrete distributions of each characteristic were established by taking visual measurements over discrete areas of PSL boards, then planing the boards down 3 mm to the next surface, then measuring again for 4 successive layers to get a thru-thickness profile. A total of 2500 measurements were taken to define the distributions of each characteristic. These distributions were then incorporated into the computer model. The upgraded model was used to predict the load/displacement behaviour of small scale PSL specimens under tension, compression and three-point bending conditions. Meanwhile, experimental tests were performed for comparison. Cumulative distribution function (CDF) curves for ultimate stress were developed for both predicted and experimental data for each test configuration. The model results were extremely accurate.
Comparing the 50th percentile, the percent error was a mere 5.7%, 2.4% and 2.0 % for tension, compression and bending, respectively. These results are extremely significant scientifically as they establish the veracity of the model under three common applications for PSL. The analysis was restricted, however to small specimens which is not representative of construction conditions. Further validation of the model on full-scale specimens (construction size) is still necessary. (Grad students = 1)
Impacts
This year's progress has completed the first specific objective of the project (successfully incorporating PSL characteristics into the model) and has taken us closer to our ultimate goal of developing a comprehensive, user-friendly computer model to simulate the mechanical properties of wood based composites. This model will aid development of new wood composites and thereby lead to better utilization of wood and wood fibre.
Publications
- No publications reported this period
|
Progress 10/01/01 to 09/30/02
Outputs
The goal of the current study is to develop a comprehensive, user-friendly computer model to simulate the mechanical properties of wood based composites. The model is an extension of an already very sophisticated approach developed by the Project Director. The existing approach has the capability to simulate stress/strain behavior for tensile, compressive or bending loading of small scale laminates with varying grain angle. The current work applies this approach to the commercial product Parallel Strand Lumber, PSL. The strength and stiffness properties of PSL in tension, compression and shear are characterized by appropriate statistical distributions. These properties are then used as direct input into the model to predict the constitutive behaviour of static bending specimens under a variety of load scenarios of the same product. The purpose of this work is to test the robustness of the current model to predict the strength behaviour of PSL. The exploratory work
will expose weaknesses in the model and provide opportunity to 'fine-tune' the approach in the attempt to make it general enough to be used for a variety of laminated composites. Grad Students = 0.
Impacts
Using a computer model to predict material strength will help alleviate the need for expensive and time consuming empirical tests for development of new wood composites. Reducing cost and time will encourage companies to investigate new products and, in the end, support the environment by reducing wood waste and optimizing the wood fiber resource.
Publications
- Clouston, P.; Lam, F. 2002. A Stochastic Plasticity Approach to Strength Modeling of Strand-based Wood Composites. Composites Science and Technology, 62 (2002), pp. 1381-1395
- Clouston, P.; Lam, F. 2001. Computational Modeling of Strand-based Wood Composites. ASCE Journal of Engineering Mechanics , 127(8), 844-851
|
|