Building Dynamic Data Centers for Fast Delivery of New Data and Data Updates

Poster at Open Repositories 2014, Helsinki, Finland, June 9-13, 2014

Scheduling dynamic dataflow graphs with model checking

With the shift towards many-core computer architectures, dataflow programming has been proposed as one potential solution for producing software that scales to a varying number of processor cores. Programming for parallel architectures is considered difficult as the current popular programming languages are inherently sequential and introducing parallelism is typically up to the programmer. Dataflow, however, is inherently parallel, describing an application as a directed graph, where nodes represent calculations and edges represent a data dependency in form of a queue. These queues are the only allowed communication between the nodes, making the dependencies between the nodes explicit and thereby also the parallelism. Once a node have the su cient inputs available, the node can, independently of any other node, perform calculations, consume inputs, and produce outputs. Data ow models have existed for several decades and have become popular for describing signal processing applications as the graph representation is a very natural representation within this eld. Digital lters are typically described with boxes and arrows also in textbooks. Data ow is also becoming more interesting in other domains, and in principle, any application working on an information stream ts the dataflow paradigm. Such applications are, among others, network protocols, cryptography, and multimedia applications. As an example, the MPEG group standardized a dataflow language called RVC-CAL to be use within reconfigurable video coding. Describing a video coder as a data ow network instead of with conventional programming languages, makes the coder more readable as it describes how the video dataflows through the different coding tools. While dataflow provides an intuitive representation for many applications, it also introduces some new problems that need to be solved in order for data ow to be more widely used. The explicit parallelism of a dataflow program is descriptive and enables an improved utilization of available processing units, however, the independent nodes also implies that some kind of scheduling is required. The need for efficient scheduling becomes even more evident when the number of nodes is larger than the number of processing units and several nodes are running concurrently on one processor core. There exist several data ow models of computation, with different trade-offs between expressiveness and analyzability. These vary from rather restricted but statically schedulable, with minimal scheduling overhead, to dynamic where each ring requires a ring rule to evaluated. The model used in this work, namely RVC-CAL, is a very expressive language, and in the general case it requires dynamic scheduling, however, the strong encapsulation of dataflow nodes enables analysis and the scheduling overhead can be reduced by using quasi-static, or piecewise static, scheduling techniques. The scheduling problem is concerned with nding the few scheduling decisions that must be run-time, while most decisions are pre-calculated. The result is then an, as small as possible, set of static schedules that are dynamically scheduled. To identify these dynamic decisions and to find the concrete schedules, this thesis shows how quasi-static scheduling can be represented as a model checking problem. This involves identifying the relevant information to generate a minimal but complete model to be used for model checking. The model must describe everything that may affect scheduling of the application while omitting everything else in order to avoid state space explosion. This kind of simplification is necessary to make the state space analysis feasible. For the model checker to nd the actual schedules, a set of scheduling strategies are de ned which are able to produce quasi-static schedulers for a wide range of applications. The results of this work show that actor composition with quasi-static scheduling can be used to transform data ow programs to t many different computer architecture with different type and number of cores. This in turn, enables dataflow to provide a more platform independent representation as one application can be fitted to a specific processor architecture without changing the actual program representation. Instead, the program representation is in the context of design space exploration optimized by the development tools to fit the target platform. This work focuses on representing the dataflow scheduling problem as a model checking problem and is implemented as part of a compiler infrastructure. The thesis also presents experimental results as evidence of the usefulness of the approach.

Laser cladding with scanning optics

Scanning optics create different types of phenomena and limitation to cladding process compared to cladding with static optics. This work concentrates on identifying and explaining the special features of laser cladding with scanning optics. Scanner optics changes cladding process energy input mechanics. Laser energy is introduced into the process through a relatively small laser spot which moves rapidly back and forth, distributing the energy to a relatively large area. The moving laser spot was noticed to cause dynamic movement in the melt pool. Due to different energy input mechanism scanner optic can make cladding process unstable if parameter selection is not done carefully. Especially laser beam intensity and scanning frequency have significant role in the process stability. The laser beam scanning frequency determines how long the laser beam affects with specific place local specific energy input. It was determined that if the scanning frequency in too low, under 40 Hz, scanned beam can start to vaporize material. The intensity in turn determines on how large package this energy is brought and if the intensity of the laser beam was too high, over 191 kW/cm2, laser beam started to vaporize material. If there was vapor formation noticed in the melt pool, the process starts to resample more laser alloying due to deep penetration of laser beam in to the substrate. Scanner optics enables more flexibility to the process than static optics. The numerical adjustment of scanning amplitude enables clad bead width adjustment. In turn scanner power modulation (where laser power is adjusted according to where the scanner is pointing) enables modification of clad bead cross-section geometry when laser power can be adjusted locally and thus affect how much laser beam melts material in each sector. Power modulation is also an important factor in terms of process stability. When a linear scanner is used, oscillating the scanning mirror causes a dwell time in scanning amplitude border area, where the scanning mirror changes the direction of movement. This can cause excessive energy input to this area which in turn can cause vaporization and process instability. This process instability can be avoided by decreasing energy in this region by power modulation. Powder feeding parameters have a significant role in terms of process stability. It was determined that with certain powder feeding parameter combinations powder cloud behavior became unstable, due to the vaporizing powder material in powder cloud. Mainly this was noticed, when either or both the scanning frequency or powder feeding gas flow was low or steep powder feeding angle was used. When powder material vaporization occurred, it created vapor flow, which prevented powder material to reach the melt pool and thus dilution increased. Also powder material vaporization was noticed to produce emission of light at wavelength range of visible light. This emission intensity was noticed to be correlated with the amount of vaporization in the powder cloud.

Modal testing and numerical modeling of the dynamic properties of layered sheet-steel structure

Recently, due to the increasing total construction and transportation cost and difficulties associated with handling massive structural components or assemblies, there has been increasing financial pressure to reduce structural weight. Furthermore, advances in material technology coupled with continuing advances in design tools and techniques have encouraged engineers to vary and combine materials, offering new opportunities to reduce the weight of mechanical structures. These new lower mass systems, however, are more susceptible to inherent imbalances, a weakness that can result in higher shock and harmonic resonances which leads to poor structural dynamic performances. The objective of this thesis is the modeling of layered sheet steel elements, to accurately predict dynamic performance. During the development of the layered sheet steel model, the numerical modeling approach, the Finite Element Analysis and the Experimental Modal Analysis are applied in building a modal model of the layered sheet steel elements. Furthermore, in view of getting a better understanding of the dynamic behavior of layered sheet steel, several binding methods have been studied to understand and demonstrate how a binding method affects the dynamic behavior of layered sheet steel elements when compared to single homogeneous steel plate. Based on the developed layered sheet steel model, the dynamic behavior of a lightweight wheel structure to be used as the structure for the stator of an outer rotor Direct-Drive Permanent Magnet Synchronous Generator designed for high-power wind turbines is studied.

Cardiorespiratory adaptations induced by aerobic training in middle-aged men: the importance of a decrease in sympathetic stimulation for the contribution of dynamic exercise tachycardia

We investigated the effects of aerobic training on the efferent autonomic control of heart rate (HR) during dynamic exercise in middle-aged men, eight of whom underwent exercise training (T) while the other seven continued their sedentary (S) life style. The training was conducted over 10 months (three 1-h sessions/week on a field track at 70-85% of the peak HR). The contribution of sympathetic and parasympathetic exercise tachycardia was determined in terms of differences in the time constant effects on the HR response obtained using a discontinuous protocol (4-min tests at 25, 50, 100 and 125 watts on a cycle ergometer), and a continuous protocol (25 watts/min until exhaustion) allowed the quantification of the parameters (anaerobic threshold, VO2 AT; peak O2 uptake, VO2 peak; power peak) that reflect oxygen transport. The results obtained for the S and the T groups were: 1) a smaller resting HR in T (66 beats/min) when compared to S (84 beats/min); 2) during exercise, a small increase in the fast tachycardia (D0-10 s) related to vagal withdrawal (P<0.05, only at 25 watts) was observed in T at all powers; at middle and higher powers a significant decrease (P<0.05 at 50, 100 and 125 watts) in the slow tachycardia (D1-4 min) related to a sympathetic-dependent mechanism was observed in T; 3) the VO2 AT (S = 1.06 and T = 1.33 l/min) and VO2 peak (S = 1.97 and T = 2.47 l/min) were higher in T (P<0.05). These results demonstrate that aerobic training can induce significant physiological adaptations in middle-aged men, mainly expressed as a decrease in the sympathetic effects on heart rate associated with an increase in oxygen transport during dynamic exercise.

Endogenous vasopressin and the central control of heart rate during dynamic exercise

The present article contains a brief review on the role of vasopressinergic projections to the nucleus tractus solitarii in the genesis of reflex bradycardia and in the modulation of heart rate control during exercise. The effects of vasopressin on exercise tachycardia are discussed on the basis of both the endogenous peptide content changes and the heart rate response changes observed during running in sedentary and trained rats. Dynamic exercise caused a specific vasopressin content increase in dorsal and ventral brainstem areas. In accordance, rats pretreated with the peptide or the V1 blocker into the nucleus tractus solitarii showed a significant potentiation or a marked blunting of the exercise tachycardia, respectively, without any change in the pressure response to exercise. It is proposed that the long-descending vasopressinergic pathway to the nucleus tractus solitarii serves as one link between the two main neural controllers of circulation, i.e., the central command and feedback control mechanisms driven by the peripheral receptors. Therefore, vasopressinergic input could contribute to the adjustment of heart rate response (and cardiac output) to the circulatory demand during exercise.

Two-phase flow measurements with spacer grid in a rod bundle geometry

Heat transfer effectiveness in nuclear rod bundles is of great importance to nuclear reactor safety and economics. An important design parameter is the Critical Heat Flux (CHF), which limits the transferred heat from the fuel to the coolant. The CHF is determined by flow behaviour, especially the turbulence created inside the fuel rod bundle. Adiabatic experiments can be used to characterize the flow behaviour separately from the heat transfer phenomena in diabatic flow. To enhance the turbulence, mixing vanes are attached to spacer grids, which hold the rods in place. The vanes either make the flow swirl around a single sub-channel or induce cross-mixing between adjacent sub-channels. In adiabatic two-phase conditions an important phenomenon that can be investigated is the effect of the spacer on canceling the lift force, which collects the small bubbles to the rod surfaces leading to decreased CHF in diabatic conditions and thus limits the reactor power. Computational Fluid Dynamics (CFD) can be used to simulate the flow numerically and to test how different spacer configurations affect the flow. Experimental data is needed to validate and verify the used CFD models. Especially the modeling of turbulence is challenging even for single-phase flow inside the complex sub-channel geometry. In two-phase flow other factors such as bubble dynamics further complicate the modeling. To investigate the spacer grid effect on two-phase flow, and to provide further experimental data for CFD validation, a series of experiments was run on an adiabatic sub-channel flow loop using a duct-type spacer grid with different configurations. Utilizing the wire-mesh sensor technology, the facility gives high resolution experimental data in both time and space. The experimental results indicate that the duct-type spacer grid is less effective in canceling the lift force effect than the egg-crate type spacer tested earlier.

Defining the keyhole modes – the effects on the weld geometry and the molten pool behaviour in high power laser welding of stainless steels

Keyhole welding, meaning that the laser beam forms a vapour cavity inside the steel, is one of the two types of laser welding processes and currently it is used in few industrial applications. Modern high power solid state lasers are becoming more used generally, but not all process fundamentals and phenomena of the process are well known and understanding of these helps to improve quality of final products. This study concentrates on the process fundamentals and the behaviour of the keyhole welding process by the means of real time high speed x-ray videography. One of the problem areas in laser welding has been mixing of the filler wire into the weld; the phenomena are explained and also one possible solution for this problem is presented in this study. The argument of this thesis is that the keyhole laser welding process has three keyhole modes that behave differently. These modes are trap, cylinder and kaleidoscope. Two of these have sub-modes, in which the keyhole behaves similarly but the molten pool changes behaviour and geometry of the resulting weld is different. X-ray videography was used to visualize the actual keyhole side view profile during the welding process. Several methods were applied to analyse and compile high speed x-ray video data to achieve a clearer image of the keyhole side view. Averaging was used to measure the keyhole side view outline, which was used to reconstruct a 3D-model of the actual keyhole. This 3D-model was taken as basis for calculation of the vapour volume inside of the keyhole for each laser parameter combination and joint geometry. Four different joint geometries were tested, partial penetration bead on plate and I-butt joint and full penetration bead on plate and I-butt joint. The comparison was performed with selected pairs and also compared all combinations together.

Ambulatory blood pressure and Doppler echocardiographic indexes of borderline hypertensive men presenting an exaggerated blood pressure response during dynamic exercise

Borderline hypertension (BH) has been associated with an exaggerated blood pressure (BP) response during laboratory stressors. However, the incidence of target organ damage in this condition and its relation to BP hyperreactivity is an unsettled issue. Thus, we assessed the Doppler echocardiographic profile of a group of BH men (N = 36) according to office BP measurements with exaggerated BP in the cycloergometric test. A group of normotensive men (NT, N = 36) with a normal BP response during the cycloergometric test was used as control. To assess vascular function and reactivity, all subjects were submitted to the cold pressor test. Before Doppler echocardiography, the BP profile of all subjects was evaluated by 24-h ambulatory BP monitoring. All subjects from the NT group presented normal monitored levels of BP. In contrast, 19 subjects from the original BH group presented normal monitored BP levels and 17 presented elevated monitored BP levels. In the NT group all Doppler echocardiographic indexes were normal. All subjects from the original BH group presented normal left ventricular mass and geometrical pattern. However, in the subjects with elevated monitored BP levels, fractional shortening was greater, isovolumetric relaxation time longer, and early to late flow velocity ratio was reduced in relation to subjects from the original BH group with normal monitored BP levels (P<0.05). These subjects also presented an exaggerated BP response during the cold pressor test. These results support the notion of an integrated pattern of cardiac and vascular adaptation during the development of hypertension.

Spectral analysis of the electromyograph of the erector spinae muscle before and after a dynamic manual load-lifting test

The aim of the present study was to assess the spectral behavior of the erector spinae muscle during isometric contractions performed before and after a dynamic manual load-lifting test carried out by the trunk in order to determine the capacity of muscle to perform this task. Nine healthy female students participated in the experiment. Their average age, height, and body mass (± SD) were 20 ± 1 years, 1.6 ± 0.03 m, and 53 ± 4 kg, respectively. The development of muscle fatigue was assessed by spectral analysis (median frequency) and root mean square with time. The test consisted of repeated bending movements from the trunk, starting from a 45º angle of flexion, with the application of approximately 15, 25 and 50% of maximum individual load, to the stand up position. The protocol used proved to be more reliable with loads exceeding 50% of the maximum for the identification of muscle fatigue by electromyography as a function of time. Most of the volunteers showed an increase in root mean square versus time on both the right (N = 7) and the left (N = 6) side, indicating a tendency to become fatigued. With respect to the changes in median frequency of the electromyographic signal, the loads used in this study had no significant effect on either the right or the left side of the erector spinae muscle at this frequency, suggesting that a higher amount and percentage of loads would produce more substantial results in the study of isotonic contractions.

Identification of anaerobic threshold using heart rate response during dynamic exercise

The objective of the present study was to characterize the heart rate (HR) patterns of healthy males using the autoregressive integrated moving average (ARIMA) model over a power range assumed to correspond to the anaerobic threshold (AT) during discontinuous dynamic exercise tests (DDET). Nine young (22.3 ± 1.57 years) and 9 middle-aged (MA) volunteers (43.2 ± 3.53 years) performed three DDET on a cycle ergometer. Protocol I: DDET in steps with progressive power increases of 10 W; protocol II: DDET using the same power values as protocol 1, but applied randomly; protocol III: continuous dynamic exercise protocol with ventilatory and metabolic measurements (10 W/min ramp power), for the measurement of ventilatory AT. HR was recorded and stored beat-to-beat during DDET, and analyzed using the ARIMA (protocols I and II). The DDET experiments showed that the median physical exercise workloads at which AT occurred were similar for protocols I and II, i.e., AT occurred between 75 W (116 bpm) and 85 W (116 bpm) for the young group and between 60 W (96 bpm) and 75 W (107 bpm) for group MA in protocols I and II, respectively; in two MA volunteers the ventilatory AT occurred at 90 W (108 bpm) and 95 W (111 bpm). This corresponded to the same power values of the positive trend in HR responses. The change in HR response using ARIMA models at submaximal dynamic exercise powers proved to be a promising approach for detecting AT in normal volunteers.

Dynamic carbon dioxide liquidization model

In literature CO 2 liquidization is well studied with steady state modeling. Steady state modeling gives an overview of the process but it doesn’t give information about process behavior during transients. In this master’s thesis three dynamic models of CO2 liquidization were made and tested. Models were straight multi-stage compression model and two compression liquid pumping models, one with and one without cold energy recovery. Models were made with Apros software, models were also used to verify that Apros is capable to model phase changes and over critical state of CO 2. Models were verified against compressor manufacturer’s data and simulation results presented in literature. From the models made in this thesis, straight compression model was found to be the most energy efficient and fastest to react to transients. Also Apros was found to be capable tool for dynamic liquidization modeling.

Dynamic linkages of real estate and stock markets in Finland

Traditionally real estate has been seen as a good diversification tool for a stock portfolio due to the lower return and volatility characteristics of real estate investments. However, the diversification benefits of a multi-asset portfolio depend on how the different asset classes co-move in the short- and long-run. As the asset classes are affected by the same macroeconomic factors, interrelationships limiting the diversification benefits could exist. This master’s thesis aims to identify such dynamic linkages in the Finnish real estate and stock markets. The results are beneficial for portfolio optimization tasks as well as for policy-making. The real estate industry can be divided into direct and securitized markets. In this thesis the direct market is depicted by the Finnish housing market index. The securitized market is proxied by the Finnish all-sectors securitized real estate index and by a European residential Real Estate Investment Trust index. The stock market is depicted by OMX Helsinki Cap index. Several macroeconomic variables are incorporated as well. The methodology of this thesis is based on the Vector Autoregressive (VAR) models. The long-run dynamic linkages are studied with Johansen’s cointegration tests and the short-run interrelationships are examined with Granger-causality tests. In addition, impulse response functions and forecast error variance decomposition analyses are used for robustness checks. The results show that long-run co-movement, or cointegration, did not exist between the housing and stock markets during the sample period. This indicates diversification benefits in the long-run. However, cointegration between the stock and securitized real estate markets was identified. This indicates limited diversification benefits and shows that the listed real estate market in Finland is not matured enough to be considered a separate market from the general stock market. Moreover, while securitized real estate was shown to cointegrate with the housing market in the long-run, the two markets are still too different in their characteristics to be used as substitutes in a multi-asset portfolio. This implies that the capital intensiveness of housing investments cannot be circumvented by investing in securitized real estate.

Optimization of a dynamic district heating power plant

The purpose of this Thesis is to find the most optimal heat recovery solution for Wärtsilä’s dynamic district heating power plant considering Germany energy markets as in Germany government pays subsidies for CHP plants in order to increase its share of domestic power production to 25 % by 2020. Different heat recovery connections have been simulated dozens to be able to determine the most efficient heat recovery connections. The purpose is also to study feasibility of different heat recovery connections in the dynamic district heating power plant in the Germany markets thus taking into consideration the day ahead electricity prices, district heating network temperatures and CHP subsidies accordingly. The auxiliary cooling, dynamical operation and cost efficiency of the power plant is also investigated.