Coursework confirmation Yes Hours supervisor spent with candidate 3 Declaration YesPhysics Extended Essay Parameter Calibration of the Voellmy-Salm Avalanche Dynamics Model Total Word Count: 3992Extended Essay Physics February 2018 Abstract Gravitational mass movements such as snow avalanches and debris flows are natural hazards. With their great destructive power, they can severely damage infrastructures and thus pose a great threat to life in mountainous regions. In order to classify areas as safe or at risk, the Voellmy-Salm model has been developed, which is capable of calculating the expected runout distance (point of furthest reach) of snow avalanches and debris flows. This model includes two friction parameters: 1 accounts for dry Coulomb-type friction and € accounts for turbulent friction. Because these parameters depend on many factors, the values of each gravitational mass movement have to be evaluated individually. This is referred to as parameter calibration. This paper includes an experiment, in which | let sand slide down a seff-built, wooden chute with four meters length and one meter width. The runout distance and flow velocity at four different inclinations is recorded. | investigate the research question: Which specific pair of values for € and p is needed to accurately calculate the runout distances of experimental sand flows at four different inclinations using the Voellmy-Salm model? In other words, | aim to find a specific value for both § and u so that the Voellmy-Salm model's calculated runout distances match the experimental runout distances. This process is called parameter calibration by back-analysis. Using trial-and-error, the best match is achieved with ) = 0.29 and € = 255ms*, resulting in a mean percentage error of below 2%. | have thus found a suitable pair of values for and , which enables the precise modeling of small-scale sand flows in my specific chute using the Voellmy-Salm model. The essay further shows that these values are consistent with other scientific research. Word Count: 286Extended Essay Physics February 2018 Table of Contents Abstract. 1. Introduction .. 2. Voellmy-Salm Model 3. Experiment 3.1 Chute.. 3.2 Methodology... 3.3 Variables. 3.3.1 Independent Variable p.. 3.3.2 Dependent Variable s 3.3.3 Controlled Variable dO. 3.3.4 Controlled Variable gs... 4. Analysis... 4.1 Velocity Analysis oR oS 1 4,2 Runout Distance Analysis .. 5. Conclusion ... 6, Discussion... 7. Evaluation .... 8, Bibliography 9. Appendix... 9.1 Example Excel Calculation, 9.2 Linear Regression .....e si 9.3 Extended Essay External Mentor Form...Extended Essay Physics February 2018 1. Introduction 1am a very passionate back country skier and avalanches have thus always fascinated me. With great respect, they are sometimes called “the white death”. Even the smallest of avalanches can bring death to careless skiers, and major ones can easily erase a whole mountain village. In order to provide safety for mountain dwellers and infrastructure such as roads and railways, itis extremely important to understand the dynamics of avalanches. Numerous avalanche dynamics models have been developed so far, of which the Voelimy-Salm model is the most popular [L1]. It is capable of calculating the expected maximum runout distance (point of furthest reach), as well as the impact pressures of avalanches. This information is then used to classify areas as "safe" or "unsafe". [L2] While studying this model, | learned that itis extremely difficult to model snow avalanches because of the complex and thus largely unknown rheological behavior of the Non- Newtonian material snow [GFM, p. 34]. In the case of the Voellmy-Salm model, the rheological behavior of snow is represented by the use of two friction parameters € and p. Their values depend on many factors such as material, flow size, and bed roughness. Yet, their values can only be determined empirically. They are calibrated to statistical data of past avalanche events, as well as to small- and large-scale experiments. In this process of back-analysis, one aims to find a configuration of § and 1 so that the model's calculations exactly fit the real data [D. Mancarella, 2010} | thus decided to perform such a back-analysis and parameter calibration of the Voellmy- Salm model by using a self-designed experiment. Since this paper had to be written before winter season, | could not use snow. However, according to definition, avalanches are gravitational mass movements that can consist of many materials. Snow avalanches, 3Extended Essay Physics February 2018 debris flows, and mudslides are physically closely related [L7]. Therefore, the Voellmy- Salm model is also a frequently used rheology in debris flow modeling [H. Y. Hussin et al., 2012]. This means that | could expect the model to be applicable to sand flows as well. In my experiment, | let sand slide down a self-constructed wooden chute to investigate the following research question: + Which specific pair of values for £ and 1 is needed to accurately calculate the runout distances of experimental sand flows at four different inclinations using the Voellmy-Salm model? 2. Voellmy-Salm Model [L5; HTL, p. 72{f.; L6; HTL, p. 76ff, [B. Salm et al., 1990] The Voellmy-Salm model was created by the Swiss engineers Dr. Adolf Voellmy and Dr. Bruno Salm. The model divides the avalanche path into a release, acceleration, and runout zone. There are two opposing forces acting on every gravitational mass movement. The accelerating force F, = sin()+m-g acts parallel to the slope with gradient @. The decelerating force F. is composed of various forces such as dry Coulomb friction, as welll as internal shear, deformation, and momentum exchange [L3] [L4]. In this physical model, the mass movement is simplified to its governing physical laws: the conservation of mass, energy, and momentum. All hese laws can be expressed by partial differential equations to describe the behavior of the medium. In order to produce theExtended Essay Physics February 2018 desired outcomes (e.g. runout distance) with this underdetermined system of equations (meaning that there are more unknowns that equations), the system needs to be “closed”. This is done by adding information about the materials rheology in the form of constitutive equations. They characterize the individual material and its macroscopic reaction to applied loads (= forces [L11]) resulting from its internal constitution [L. E. Malvern, 1969}. Until today, the exact constitutive equations of snow and debris are largely unknown and models like that of Voellmy-Salm are thus based on grossly simplified concepts. More precisely, the Voellmy-Salm model uses the constitutive equation known as the 2- parametric-Voellmy-friction model or Voellmy-rheology. As the name says, the decelerating friction force F, is modeled as an interaction of two components [K. Hutter, 1989] [L8): 1. The dry Coulomb-type basal friction. It is scaled with the normal force Fwand has friction parameter p (unitless). This friction dominates when the mass movement mainly consists of larger tumbling and sliding blocks and has the form of a granular solid [C. Ancey et al., 2004]. This is the case at lower flow speeds [L12], especially in the runout zone. 2. The turbulent hydrodynamic friction has parameter § [m/s*]. As the mass movement gains speed, the momentum of individual particles increases. Therefore, they rather break up than consolidate to larger blocks. As a result, the movement gains the properties of a granular fluid, which is then subject to turbulent drag [L12]. This drag is produced by obstacles, making § dependent on the avalanche path's roughness and canalization. Table 1 lists the variables and parameters used in the calculations.Extended Essay Physics February 2018 s Runout distance [m] % flow rate / intial flow rate [| By. Bp Flow width in release zone / at point P [m] dy Initial low height prior to release [m] dy, d, Flow height at point P/in runout zone [m] Vo. BPs Velocity when leaving release zone / at point P [=] m Mass of flow [kg] g Gravitational constant (9.81 a) 0, >. s Mean gradient in release zone / Point P/ runout zone [°] OK Critical gradient F] (auxiliary variable) mean avalanche v velocity |] iu dry-Coulomb friction [unitiess}] 5 Turbulent friction ["] Table 1: Variables and parameters of the Voellmy-Salm model The mass balance equation states that the flow rate Q stays constant during the descend. Itis expressed as Q = Qo = By“ do“ vo a By is obtained from observations in the field or on the map. In order to find vo, the momentum equation fe gor, mapa fa is solved. Implementing the Voellmy-rheology, this becomesExtended Essay Physics February 2018 dv gv? me = mg sin(@) — mg cos(p) ~ FG @) The middle term represents dry friction with parameter jx and Fy = mg cos(‘p). The right term stands for turbulent, which is proportional to velocity squared. The flow’s friction increases as 1 increases and § decreases. Voelimy-Salm's model is rooted in the hydraulic theory. Therefore, "the avalanche is modelled as a fluid which accelerates very quickly from rest to a steady, terminal velocity’ IR. Perla et al., 1980]. The velocities calculated with this model are thus regarded constant, with 2 = 0. Equation 2 thereby becomes 2 Eq = sine) — neos(y) by canceling out mg. The velocity vp can be calculated with this equation: v= Hcos(o)) At this point we observe that the results of this model strongly depend on parameters do,wyand 8 Now, point P is determined. It is the point with critical gradient , = arctan(j). It marks the beginning of the runout zone with gradient «, f M = 8289. With this result, | carry out a x?-test on the goodness-of-fit of a straight line to our data. Since | started with n = 4 means and have determined r = 2 unknown parameters, the degree of freedom is n — r = 2. With a significance level of 0.01, we find the 32Extended Essay Physics February 2018 corresponding critical value in Table 9 to be 9.21. As M < 9.21, there is thus no reason to reject the hypothesis of a straight line. Critical Vals ofthe x? ; Mert peed pete ene Table 9: Critical values of the x*-distribution [L10] 33,Extended Essay Physics February 2018 9.3 Extended Essay External Mentor Form ‘To be read by the external mentor before the DP student begins his/her research opportunity under external guidance This form mus be signed by the external mentor when the projects completed Information about the extended estay process and external mentoring ‘Tneextended essays a rigorous ad challenging independent research project carried ot by students during the Diploma Programme Students conducting an extended esi ae personally responsible for ‘every aspect ofthe research process. They are required to: 1 develop their own research question develop their own appropriate and ethical research methods, which must comply with 1B polices and regulations Collect and process their own data ‘show critical understanding ofthe approaches wed + demonstrate academic meg. Each student has schoo supervisor to guide them through the whole process. The rola he external ‘mentor sto Inform the student of general research opportunities and to teach and oversee the sae rid appropriate useof techniques. ‘Toavoid compromising the integrity ofthe process, the external mentor shoud nt 1 provide the student with 2 research question or focused topic provide ready-to-use techniques and protocols Provide advanced or extensive technician support Provide ready-to-use data analysis tools 1 ead, edit or comment on any writen work. 34Extended Essay Physics, February 2018 ‘Tobe signed by the external mentor: ‘Ihave been made aware of the requirements ofthe IB Diploma Programme extended essay and ofthe ‘esponsibiliies of an external mentor Ihave ated asthe mentor fr: — ame of student Comment Name: Hearne signature [Email address: Date: Qh OB ZODPDF forms are not compatile with the Google Chrome POF viewer plug-in. Chrome users shoul save the form, then regpen and complete with Adobe reader EE/RPPF SP Wrernational Baccalaureate Baccalauréat international For use from MayiNovember 2018 OF sschiterto iternacional Candidate personal code: Extended essay - Reflections on planning and progress form Candidate: This form is to be completed by the candidate during the course and completion oftheir EE. This document records reflections on your planning and progress, and the nature of your discussions with your supervisor. You must undertake three formal reflection sessions with your supervisor: The first formal reflection session should focus on your intial ideas and how you plan to undertake your research the interim reflection session is once a significant amount of your research has been completed, ‘and the final session will bein the form of a viva voce once you have completed and handed in your EE. This document acts as a record in supporting the authenticity of your work. The three reflections combined must amount to no more than 500 words. ‘The completion of this form is a mandatory requirement of the EE for first assessment May 2018. It must be submitted together with the completed EE for assessment under Criterion E. ‘Supervisor: You must have three reflection sessions with each candidate, one early on in the process, an interim meeting and then the final viva voce. Other check-in sessions are permitted but do not need to be recorded on this sheet. After each reflection ‘session candidates must record their reflections and as the supervisor you must sign and date this form. First reflection session Candidate comments: [Being a passionate skier, avalanches have always fascinated me. When doing the initial research into the physical |description of avalanches, | was overwhelmed by the sheer complexity and unfamiliar physics of this topic. it became clear Ithat external help was essential. After contacting the Institute for Snow and Avalanche Research SLF, | managed to get in ltouch with who was willing to introduce me to the Voellmy-Salm avalanche calculation madel. This mentoring Itook one day. He only taught me the Voellmy-Salm model but did not provide any help regarding the research question or lexperiment. Inspired by the snow chute of the SLF, | decided to build my own, smaller chute. Unfortunately, it was still summer and I thus |could not use snow. Therefore, my initial research objective was to compare the flow behavior of experimental sand flows to |snow avalanches, which are calculated with the Voellmy-Salm model. However, in the initial reflection session, my [supervisor reminded me of the nature of physical models with their significant limitations. | thus realized that it would not be lappropriate to draw conclusions from a comparison between an experiment and a model, since both of them represent only la simplification and abstraction of the real phenomena. At this point, | was unsure whether a sand-based experiment is even |compatible with the snow avalanche theory | have learnt so far. Date: [06.12.2017 ‘Supervisor inital: sea‘caret trations, Page2T [heist eras! EE/RPPF Dd sora scr Interim reflection Candidate comments [Having become more familar with the Voelimy-Salm model, | earnt that his model is also used for debris flow modeling, lwhich bear a strong resemblance to sand flows. Therefore, | could now focus my research on modeling my experimental lsand flows with the Voellmy-Salm model. The new research question, which addresses the calibration of the model's friction lparameters, is now precise enough to be treated within the word count. IMy supervisor and | reviewed the sources | referenced in my first draft. While most of the them were very reliable (ike the ISLF website and all books and studies), | also used some less trustworthy sources, such as Wikipedia. They were then replaced with accepted sources. My supervisor also lent me a book by A. Tipler and Gene Mosca, from where | took the statistical methods utilized in the essay. Date: [09.15.2017 ‘Supervisor ina: Final reflection - Viva voce Candidate comments: [The process worked out as anticipated. Suitable values for the fiction parameters could be found. | was especially proud of |my chute. | was glad that | put so much though into its stable, durable, and modular design, which enabled many lexperimental runs in a short amount of time. When reading other, sometimes lengthy, research papers, | had to Icistinguishing between necessary and unnecessary information, which eventually improved my research skills. Furthermore, I got personally very interested in the area of rheology and continuum mechanics and | would like to learn more about this at University. Date: [01252078 Supervisor initials: meeeEE/RPPF Ox Popes [heist eros! ‘Supervisor comments: ‘Supervisor: By submiting this candidate work for assessment, you are taking responsibly forts authenticity. No piace of candidate work ‘should be uploaded/submited to the e-Coursework system fits authentic isin doubt or if contradictory comments are added to this form. If your tert in the box below raises any doubt on the authenticity ofthe work, this component will not be assessed. [A passionate skier, the candidate began to take an interest in the physics of avalanches years ago. [The candidate tad to tackle the following issues: Neither hydrodynamics, elasticity theory, nor the so-called Voellmy-Salm Imodel form part of the regular Physics curriculum and hence aren't covered in class, which meant that the candidate had to Jobtain knowiedge in this area without the help of his supervisor. It was the candidate himself who came up with the novel idea of using sand avalanches as a model for snow avalanches. [The candidate procured and arranged for all the necessary materials and technical equipment on his own, |The candidate proved to be a highly critical thinker who challenged existing norms in the area and, when necessary, [successfully developed his own scientific criteria for the project.