Summer intern dives deep into someone else’s code

In his spare time, Saul Amster likes to program. He’s currently working on a project to turn a tablet into a magic mirror. Yes, like Snow White’s evil stepmother (“Mirror, mirror on the wall…”), except imagine asking the mirror for the day’s forecast or the score of last night’s game. “Programming is an interesting hobby,” he says. “It’s basically free. All you need is a computer. Other hobbies require you to keep sinking money into them.”

This summer, Saul turned his programming hobby into an internship at the Concord Consortium. But while he’s used other external software libraries before, he had to teach himself to work with other people’s preexisting code, plus learn the push and pull requests of contributing code on GitHub. And although he was new to the code base underlying the Seismic Explorer software, which displays earthquakes and volcanoes worldwide using real-time data from the USGS, he didn’t let that stop him from jumping in. In fact, he’s enhanced an existing feature by redrawing the plate boundaries to make them more noticeable and added a new feature that shows arrows to display the movement of the tectonic plates. He’s now reworking how the animation is done in the model.

“I have been really impressed with Saul,” says Amy Pallant, Principal Investigator of the Geological Models for Explorations of Dynamic Earth (GEODE) project, which developed Seismic Explorer. “He has been able to add new data into the model, think about the user experience, and help me make decisions about layout, design, and data representation. His vast experience with programming, computer games and educational environments meant that I could learn from him, too.”

Saul is sure this first experience working with someone else’s code will serve him well when he heads off to Ithaca College as a freshman computer science major. “This has been super helpful for classes and for future jobs,” he says. He’s not at all worried about his freshman Java course, since he has already learned the language. It’s one of his favorites, along with C#, which he uses in his videogame programming.

Saul is excited about some high-end virtual reality gear he spotted in the computer department at Ithaca. He’s already made some small VR apps for the Google cardboard, and he’s looking forward to research opportunities. So along with clothes and toothpaste, he’s packing his laptop and external graphics box—with better cooling and more power, it’s perfect for developing (and playing) games, and getting his homework done, of course.


Polish researchers independently validated Energy3D with Building Energy Simulation Test (BESTEST)

Fig. 1: BESTEST600 test case
Fig. 2: Comparison of Energy3D results with those of other simulation tools
The Building Energy Simulation Test (BESTEST) is a test developed by the International Energy Agency for evaluating various building energy simulation tools, such as EnergyPlus, BLAST, DOE2, COMFIE, ESP-r, SERIRES, S3PAS, TASE, HOT2000, and TRNSYS. The methodology is based on a combination of empirical validation, analytical verification, and comparative analysis techniques. A method was developed to systematically test whole building energy simulation programs. Geometrically simple cases, such as cases BESTEST600 to 650, are used to test the ability of a subject program to model effects such as thermal mass, direct solar gain windows, shading devices, infiltration, internal heat gain, sunspaces, earth coupling, and setback thermostat control. The BESTEST procedure has been used by most building simulation software developers as part of their standard quality control program. More information about BESTEST can be found at the U.S. Department of Energy's website.

Prof. Dr. Robert Gajewski, Head of Division of Computing in Civil Engineering, Faculty of Civil Engineering, Warsaw University of Technology, and his student Paweł Pieniążek recently used BESTEST600-630 test case (Figure 1) to evaluate the quality of Energy3D's predictions of heating and cooling costs of buildings. By comparing Energy3D's results with those from major building energy simulation tools (Figure 2), they concluded that, "[Energy3D] proved to be an excellent tool for qualitative and quantitative analysis of buildings. Such a program can be an excellent part of a computer supported design environment which takes into account also energy considerations."

Their paper was published here.

Mechanical design and paper crafting combine in Paper Mechatronics

How can you make a cardboard owl that flaps its wings? Or a paper flower that blooms? With funding from the National Science Foundation, we are working with the University of Colorado’s Craft Technology Lab and the Children’s Creativity Museum in San Francisco to study and enhance the engineering education potential of Paper Mechatronics, an innovative educational technology genre that mixes creative papercrafts, mechanical design, and computational thinking. Soon, young learners will be designing real and fantastical paper inventions of their own imagination and animate them with mechanical motions.

The new two-year project builds off an earlier project by Principal Investigators Sherry Hsi and Michael Eisenberg, which prototyped several Paper Mechatronics design projects, organized activity formats, and piloted the various design elements with children and adults to determine which worked best to inspire learning and teach design. These included a custom software design tool, simple hardware modules, cardboard electronics, sample workshop formats, and project ideas. Early Paper Mechatronics activities—from a percussion workshop to a cereal hackathon and a Robot Petting Zoo—showed encouraging results with after school youth (ages 12-18) and museum visitors.

Mechanical duck designed with Paper Mechatronics.

Robot Petting Zoo.

Paper Mechatronics engaged participants in key engineering design practices (design, build, test), though learners were challenged by translating their visions into mechanical actions. So, to support designers who had no electronics or computer-aided design background and limited computer programming experience, Ph.D. student HyunJoo Oh designed FoldMecha, which generates paper-based templates for a number of design parameters such as shape, size, and type of motor movements that can be cut out with a paper or laser cutter.

 The new project will expand and improve this early Paper Mechatronics design software for modeling mechanical components and movements and create a new Paper Mechatronics kit with instructional resources, electronically enhanced crafting materials, low-cost microcontrollers and accessories, and custom design software.

Our research goal is to explore how to support novice designers in learning from the Paper Mechatronics kit and study how youth develop adaptive expertise, including knowledge-seeking, resourcefulness, confidence, and persistence. We’ll research how on-ramps to engineering design activities like engaging in paper mechatronic design activities help youth develop adaptive expertise and what types of instructional resources and scaffolding are most useful in supporting learners to be creative in engineering design.

Introducing summer intern, data science major Maya Haigis

Before interning with senior scientist Charles Xie this summer, Maya Haigis had no idea how many solar panel manufacturers there are—“There’s a ton!”

A data science major at the University of Rochester, Maya put her analytic skills to work at the Concord Consortium collecting data on solar panels (dimensions, weight, maximum wattage, etc.) and designed a panda solar power plant with Energy3D, an engineering design and simulation tool for renewable energy and energy efficiency. She used Energy3D to create a power plant in the shape of a bald eagle, too.

“Charles heard about the giant panda power plant in Datong, China, in the news, and asked me to replicate it in Energy3D.” Maya says, “It was a good introduction to the features of Energy3D. Charles suggested I do something relevant to the U.S.—like our own national symbol! It was fun imagining flying across the country and seeing a giant bald eagle out of the window instead of the generic rectangles or circles of traditional solar farms.”

She also worked with the Energy3D team modeling local schools and other community buildings for the Solarize Your World curriculum they are designing.

“Maya is a real genius in 3D modeling,” said Charles. “I didn’t expect her to come up with sophisticated 3D structures within a couple of hours with a piece of software that she had never used before. But she did it elegantly. It is remarkable that she has created scores of highly accurate 3D models for school buildings with incredible details.”

Bald eagle solar power in Energy3D (left) and close-up of bald eagle (right).

As a sophomore, Maya is currently on the same path as computer science students, but her curriculum path will soon diverge with a focus on data mining and database systems plus more statistics. She’s always been “a math person,“ she says, but credits her high school AP statistics teacher’s enthusiasm for data and statistics for consolidating her interest.

At the University of Rochester she’s already taken courses in Java, data structures and algorithms, discrete math, calculus, and linear algebra with differential equations. “All data is interesting,” she says, but notes sports stats are particularly fascinating. No surprise, since Maya is a student athlete who plays field hockey at the Division III school where her schedule includes practice six days a week.

She notes, “My brother and I used to have a collection of baseball cards and I would try to memorize the stats of my favorite players. It’s a bit ironic because before games, coaches always say that once you step onto the field, the statistics don’t mean anything and what matters is which team plays the hardest, but I still look through other team stats.”

Recently, Maya had a pivotal experience. She spent half a day at Pfizer working with a business analyst, who serves as a connection between scientists and programmers. “The business analyst would explain to the scientists what the data meant,” she explains. “And if the scientists wanted their data displayed in a certain way, she would talk to the programmers.” Maya can imagine filling a similar liaison role working as data scientist, though she also admits, “I’m not exactly sure what I want to do after college, but I’m looking forward to the data science courses at Rochester, and I’m excited to see what opportunities will arise with big data!”

Modeling parabolic dish Stirling engines in Energy3D

Fig. 1: A parabolic dish Stirling engine
Fig. 2: The Tooele Army Depot solar project in Utah
A parabolic dish Stirling engine is a concentrated solar power (CSP) generating system that consists of a stand-alone parabolic dish reflector focusing sunlight onto a receiver positioned at the parabolic dish's focal point. The dish tracks the sun along two axes to ensure that it always faces the sun for the maximal input (for photovoltaic solar panels, this type of tracker is typically known as dual-axis azimuth-altitude tracker, or AADAT). The working fluid in the receiver is heated to 250–700 °C and then used by a Stirling engine to generate power. A Stirling engine is a heat engine that operates by cyclic compression and expansion of air or other gas (the working fluid) at different temperatures, such that there is a net conversion of thermal energy to mechanical work. The amazing Stirling engine was invented 201 years ago(!). You can see an infrared view of a Stirling engine at work in a blog article I posted early last year.

Although parabolic dish systems have not been deployed at a large scale -- compared with its parabolic trough cousin and possibly due to the same reason that AADAT is not popular in photovoltaic solar farms because of its higher installation and maintenance costs, they nonetheless provide solar-to-electric efficiency above 30%, higher than any photovoltaic solar panel in the market as of 2017.

In Version 7.2.2 of Energy3D, I have added the modeling capabilities for designing and analyzing parabolic dish engines (Figure 1). Figure 2 shows an Energy3D model of the Tooele Army Depot project in Utah. The solar power plant consists of 429 dishes, each having an aperture area of 35 square meters and outputting 3.5 kW of power.

Fig. 3: All four types of real-world CSP projects modeled in Energy3D
With this new addition, all four types of main CSP technologies -- solar towers, linear Fresnel reflectors, parabolic troughs, and parabolic dishes, have been supported in Energy3D (Figure 3). Together with its advancing ability to model photovoltaic solar power, these new features have made Energy3D one of the most comprehensive and powerful solar design and simulation software tools in the world, delivering my promise made about a year ago to model all major solar power engineering solutions in Energy3D.

An afterthought: We can regard a power tower as a large Fresnel version of a parabolic dish and the compact linear Fresnel reflectors as a large Fresnel version of a parabolic trough. Hence, all four concentrated solar power solutions are based on parabolic reflection, but with different nonimaging optical designs that strike the balance between cost and efficiency.

My Daughter Heard About an Earthquake. How Do I Explain It?

Earthquakes occur worldwide daily, and their aftereffects vary widely, from minimal to devastating. From California to the Mediterranean, some communities live with the threat and consequences of earthquakes and their aftershocks on a regular basis. Understanding what causes an earthquake is not easy. How is it possible to visualize monumental slabs of Earth moving? And why do we need to?

On June 12, 2017, newspapers worldwide reported on a 6.3 magnitude earthquake south of the island of Lesbos, Greece (off the western coast of Turkey). The quake caused widespread structural damage as well as loss of life, and it drew considerable attention, in part, because of the large number of migrants on Lesbos. How to house and care for the affected migrants and residents became a major international challenge.

But according to the USGS, the earthquake was “the result of normal faulting in the shallow crust.” The Lesbos quake was traumatic, but not unexpected. Greece and Turkey are particularly earthquake prone because they are on active fault lines. The Mediterranean region is seismically active due to the convergence of the African plate to the south with the Eurasian plate to the north. The African plate is subducting beneath the Eurasian plate at a place called the Hellenic Trench.

That’s a lot to understand, let alone visualize. When a seismic event occurs, how can a teacher explain such monumental movements of the Earth to middle school students? Typically, it’s been done with drawings and detailed descriptions, such as the excellent resources available from the USGS. But earthquakes and other geologic events are about movement, happening far out of sight. The Concord Consortium’s GEODE project is creating a way to visualize the Earth’s movements using an interactive, dynamic computer model of tectonic plates.

Another GEODE model — the Seismic Explorer — allows users to see the pattern of earthquakes worldwide, including their magnitude and depth. The Lesbos quake and many others, as well as towns and cities, are visible.

The GEODE project is still researching and developing the best ways for kids to learn about Earth’s big movements. But why is it important? Because the consequences of these movements can crumble buildings and cause loss of life. Understanding patterns of Earth’s movement may help lead to better forecasting, preparedness, and response.

Thermal imaging as a universal indicator of chemical reactions: An example of acid-base titration

Fig. 1: NaOH-HCl titration
Funded by the National Science Foundation and in collaboration with Prof. Dunwei Wang's lab at the Department of Chemistry, Boston College, we are exploring the feasibility of using thermal imaging as a universal indicator of chemical reactions. The central tenet is that, as all chemical reactions absorb or release thermal energy (endothermic or exothermic), we can infer certain information from the time evolution and spatial distribution of the temperature field.

To prove the concept, we first chose titration, a common laboratory method of quantitative chemical analysis that is used to determine the unknown concentration of an identified analyte, as a beginning example. A reagent, called the titrant, is prepared as a standard solution. A known concentration and volume of titrant reacts with a solution of analyte to determine its concentration.

The experiment we did today was an acid-base titration. An acid–base titration is the determination of the concentration of an acid or base by exactly neutralizing the acid or base with a base or acid of known concentration. Such a titration is typically done with a burette that drops titrant into an Erlenmeyer flask containing the analyte. A pH indicator is used to determine whether the equivalence point has been reached. The pH indicator usually depends on the analyte and the titrant. But a differential thermal analysis based on infrared imaging may provide a universal indicator as the technique depends only on the heat of reaction and thermal energy is universal.

Fig. 2: The dish-array titration revealed by FLIR ONE
Figures 1 and 2 in this article show the results of the NaOH+HCl titration, taken using a FLIR ONE thermal camera attached to my iPhone 6. A solution of 10% NaOH was prepared as the analyte of "unknown" concentration and 1%, 3%, 5%, 7%, 10%, 12%, 15%, 18%, and 20% HCl were used as the titrant. The experiment was conducted with a 3×3 array of Petri dishes. Hence, we call this setup as dish-array titration. Preliminary results of this first experiment appeared to be encouraging, but we have to be cautious as the dissolving of HCl after the acid-base neutralization completes can also release a significant amount of heat. How to separate the thermal signatures of reaction and dissolving requires some further thinking.

Chinese translation of SageModeler systems dynamics modeling tool

In June, Professor Silvia Wen-Yu Lee and her team at the National Changhua University of Education in Central Taiwan offered a 10-hour modeling curriculum to approximately 100 seventh grade students. Students used a new Chinese language version of SageModeler to model the relationship between marine biology and human activity in a unit about environmental conservation.

SageModeler is a free, web-based systems dynamics modeling tool for middle and high school students to construct dynamic models. SageModeler is being developed by the Building Models project, a collaboration between the Concord Consortium and the CREATE for STEM Institute at Michigan State University (MSU).

Professor Lee met Joe Krajcik, one of the lead writers of the Next Generation Science Standards and Principal Investigator (PI) of the Building Models project at MSU, where she had served as a visiting professor in 2014. Dan Damelin is the project’s PI at the Concord Consortium. Thanks to this fortuitous collaboration, Lee and her team translated SageModeler into Chinese, and her students are now taking advantage of this easy-to-use tool to create dynamic systems models.

Students building models with SageModeler. 

“The students learned how to draw models instantly after a brief demonstration,” Lee noted. “Our teachers were amazed by the students’ level of engagement and by the students’ attention to the relationships when they are working together on the SageModeler. ” Professor Lee and her colleagues at the National Changhua University of Education hope to understand how the students develop competencies in model building and whether they develop clear understandings of the causal and dynamic relationships in marine biology and human activity (fishing) through modeling.

Sample student model from a seventh grade Taiwanese student.

You can build your own model in five easy steps.

  1. Open SageModeler (in English or Chinese)

  2. Add variables to the canvas First, brainstorm factors that affect marine biology. What contributes to it and what is affected by it? Now, add images for each variable to the canvas.

  3. Link variables and set relationships Draw links from one variable to another and select from a menu to set the relationships between those variables. By using words and pictures of graphs, students can define the underlying equations that will be used to run the model.

  4. Run the model Open the simulation controls and run the model to collect data. Adjust the model settings to see how changing the variables affects the outcome. Does the model output data make sense? Does it match real-world data? Are the relationships between variables set up appropriately?

  5. Revise and expand your model Revise your model to better match the phenomenon you are modeling. For example, you may want to add more variables. As you continue to ask new questions, you can revise your model and deepen your understanding of the system.

We are currently working on additional internationalization efforts, including Turkish and Spanish translations. Interested in learning more or contributing a translation? Contact us.

Analyzing the linear Fresnel reflectors of the Sundt solar power plant in Tucson

Fig. 1: The Sundt solar power plant in Tucson, AZ
Fig. 2: Visualization of incident and reflecting light beams
Tucson Electric Power (TEP) and AREVA Solar constructed a 5 MW compact linear Fresnel reflector (CLFR) solar steam generator at TEP’s H. Wilson Sundt Generating Station -- not far from the famous Pima Air and Space Museum. The land-efficient, cost-effective CLFR technology uses rows of flat mirrors to reflect sunlight onto a linear absorber tube, in which water flows through, mounted above the mirror field. The concentrated sunlight boils the water in the tube, generating high-pressure, superheated steam for the Sundt Generating Station. The Sundt CLFR array is relatively small, so I chose it as an example to demonstrate how Energy3D can be used to design, simulate, and analyze this type of solar power plant. This article will show you how various analytic tools built in Energy3D can be used to understand a design principle and evaluate a design choice.

Fig. 3: Snapshots
One of the "strange" things that I noticed from the Google Maps of the power station (the right image in Figure 1) is that the absorber tube stretches out a bit at the northern edge of the reflector assemblies, whereas it doesn't at the southern edge. The reason that the absorber tube was designed in such a way becomes evident when we turn on the light beam visualization in Energy3D (Figure 2). As the sun rays tend to come from the south in the northern hemisphere, the focal point on the absorber tube shifts towards the north. During most days of the year, the shift decreases when the sun rises from the east to the zenith position at noon and increases when the sun lowers as it sets to the west. This shift would have resulted in what I call the edge losses if the absorber tube had not extended to the north to allow for the capture of some of the light energy bounced off the reflectors near the northern edge. This biased shift becomes less necessary for sites closer to the equator.

Energy3D has a way to "run the sun" for the selected day, creating a nice animation that shows exactly how the reflectors turn to bend the sun rays to the absorber pipe above them. Figure 3 shows five snapshots of the reflector array at 6am, 9am, 12pm, 3pm, and 6pm, respectively, on June 22 (the longest day of the year).

As we run the radiation simulation, the shadowing and blocking losses of the reflectors can be vividly visualized with the heat map (Figure 4). Unlike the heat maps for photovoltaic solar panels that show all the solar energy that hits them, the heat maps for reflectors show only the reflected portion (you can choose to show all the incident energy as well, but that is not the default).

There are several design parameters you can explore with Energy3D, such as the inter-row spacing between adjacent rows of reflectors. One of the key questions for CLFR design is: At what height should the absorber tube be installed? We can imagine that a taller absorber is more favorable as it reduces shadowing and blocking losses. The problem, however, is that, the taller the absorber is, the more it costs to build and maintain. It is probably also not very safe if it stands too tall without sufficient reinforcements. So let's do a simulation to get in the ballpark. Figure 5 shows the relationship between the daily output and the absorber height. As you can see, at six meters tall, the performance of the CLFR array is severely limited. As the absorber is elevated, the output increases but the relative gain decreases. Based on the graph, I would probably choose a value around 24 meters if I were the designer.
Fig. 4: Heat map visualization

An interesting pattern to notice from Figure 5 is a plateau (even a slight dip) around noon in the case of 6, 12, and 18 meters, as opposed to the cases of 24 and 30 meters in which the output clearly peaks at noon. The disappearance of the plateau or dip in the middle of the output curve indicates that the output of the array is probably approaching the limit.

Fig. 5: Daily output vs. absorber height
If the height of the absorber is constrained, another way to boost the output is to increase the inter-row distance gradually as the row moves away from the absorber position. But this will require more land. Engineers are always confronted with this kind of trade-offs. Exactly which solution is the optimal depends on comprehensive analysis of the specific case. This level of analysis used to be a professional's job, but with Energy3D, anyone can do it now.

Why dragons?

Breeding virtual dragons is all in a day’s work in biology classrooms using Geniverse, our free, web-based genetics software. Although Geniverse is a game-like environment, it’s far more than child’s play. Indeed, students dive into genetics on a quest to heal a beloved dragon. Students use a model species (drakes) to explore the fundamental mechanisms of heredity and genetic diseases and get a taste of careers in genetics. (Drakes are essentially a smaller version of a dragon, and are a model species in much the same way as the mouse is a model species for human genetic disease.)

But why did we choose dragons and drakes? To start, they are just plain fun! And since they’re mythical, we can bring together into one animal any and all real-world genes we’d like to teach with—without having to be restricted to a specific species’ genome. So, while our dragons and drakes are fantastical, their genes are very much real, gathered from mice, fruit flies, lizards, and other organisms we study in laboratories all over the world. When students learn genetics with Geniverse, they’ll encounter the genes again, should they venture into a real genetics lab later in life.

Students begin their Geniverse adventure as a student in the Drake Breeder’s Guild, where they move through four levels of progressively more difficult genetics challenges and unlock new chapters of the narrative. Try Geniverse now and learn how fun (and educational) dragons can be!