Geri Dön

Türbülanslı açık kanal akışlarında türbülans dağıtıcı sistemler ve kesit optimizasyonu

Turbulence eliminating systems and cross-section optimization in turbulent open channel flows

PDF İndir

  1. Tez No: 529092
  2. Yazar: CEMAL HASAN CAN AKSOY
  3. Danışmanlar: PROF. DR. ABDİ KÜKNER
  4. Tez Türü: Yüksek Lisans
  5. Konular: Mühendislik Bilimleri, Engineering Sciences
  6. Anahtar Kelimeler: Belirtilmemiş.
  7. Yıl: 2019
  8. Dil: Türkçe
  9. Üniversite: İstanbul Teknik Üniversitesi
  10. Enstitü: Fen Bilimleri Enstitüsü
  11. Ana Bilim Dalı: Gemi ve Deniz Teknoloji Mühendisliği Ana Bilim Dalı
  12. Bilim Dalı: Gemi ve Deniz Teknolojisi Mühendisliği Bilim Dalı
  13. Sayfa Sayısı: 209

Özet

Açık kanal akışlarıyla ilgili birçok çalışma olmasına rağmen, türbülansın neden olduğu çevrintilerin giderilmesi, azaltılması ya da başka bir alana taşınması amacıyla yapılan çalışmalar çok sayıda değildir. Bu amaçla ilk olarak açık kanal akışları detaylı olarak incelenmiştir. Açık kanal akışları ile ilgili olarak yapılan çalışmaların genellikle hidrolik bilim dalıyla ilgili olduğu bilinmektedir. Açık kanal akışlarıyla türbülans konusu beraber olarak incelendiğinde dahi yeterli görülebilecek çalışmalara ulaşılamamıştır. Açık kanal akışlarının incelenmesinden sonra türbülans modellerinin incelemesi yapılmıştır. Hesaplamalı akışkanlar dinamiği programlarından özellikle Fluent programının kullandığı türbülans modellerinin altyapısı gözden geçirilmiştir. Buradaki amaç yapılacak olan analiz çalışmalarında kullanılabilecek potansiyel türbülans modellerinin belirlenmesi üzerinedir. Türbülans modellerinin incelenmesi sonucunda dört farklı türbülans modeli seçilmiştir. Seçilen türbülans modelleri anahat k-ω modeli, SST k-ω modeli, gerçekleşebilir k-ε modeli ve Reynolds gerilme modelidir. Açık kanal akışlarının ve türbülans modellerinin incelemesine paralel olarak deneysel çalışmalar yürütülmüştür. Deneysel çalışmalarda iki farklı debide dört farklı parça üzerinden sekiz farklı akışın profili gözlemlenmiş, her akıştan üç farklı noktada yükseklik ölçümleri alınmıştır. Deney çalışmalarında kullanılan açık kanal yapılarının kesit tipleri ya da eğrilik oranları birbirlerinden farklıdır. Yapılan deneysel çalışmalardaki açık kanal akışları Fluent programında modellenerek seçilen dört farklı türbülans modeliyle analizleri yapılmıştır. Deneysel çalışmalarda gözlemlenen akış profili ile ölçülen değerler analiz sonuçlarıyla karşılaştırılarak en yakın sonucu veren SST k-ω türbülans modeli ana model olarak seçilmiştir. Ana model seçilmesindeki amaç yapılacak olan analiz çalışmalarında kullanılmak üzere uygun modelin bulunması üzerinedir. Ana model seçildikten sonra çevrintileri önleyebilecek sistemlerin analizleri yapılmıştır. İlk olarak türbülansın yüksek olduğu kısma katı bir malzeme konularak türbülanstaki değişim gözlenmiştir. İkinci sistem olarak ta türbülansın yapısını değiştirmesi amacıyla türbülansın yüksek olduğu kısımda analizlerde hava kabarcıkları oluşturulmuştur. Bu çalışmaların ardından dairesel, eliptik, dikdörtgen ve karmaşık geometrideki kanal kesitleri belirli eğimlerde ve dönüş açılarında analiz edilerek akış profili gözlemlenmiş, kesitlerle ve akışla ilgili bilgiler edinilmiştir.

Özet (Çeviri)

Although there are many studies on open channel flows, there are not sufficient number of studies to eliminate, reduce, or transport turbulence-induced disorders. For this purpose, open channel flows were researched in detail. It is known that studies on open channel flows are generally related to hydraulic science. These studies which are related to hydraulic science generally include the natural flow of rivers, sluice flows and weir flows. Such open channel structures do not have a geometrical cross-section normally have a natural cross-sectional form. At the same time, the roughness changes within the channel. Even if open channel flows and turbulence were examined together, the studies that could be considered adequate were not reached. When researching subjects about open channel flows, the types of open channel flows have been investigated in many respects. According to the size of the speed, critical velocity, change of the velocity, movement of liquid particles, viscosity of the fluid, type of fluid. In addition, the parameters related to open canal cross-sections were studied. Therefore Manning formula examined in detail. After researhing the open channel flows, it is investigated how the flow equations are solved numerically, and the turbulence models used in these solutions are researched in detail. The turbulence kinetic energy, energy loss rate, turbulence density, Reynolds decompisition and turbulence length scales were researched in detail. Firstly, direct numerical simulation method has been researched and it is included in the study. In particular, the reason why this method yields more accurate results, what is the relationship between the integral length scale, sub-equations and the Kolmogorov length scale were fully researched. The reasons for the high cost of calculation for direct numerical simulation were also investigated. By means of these studies, it was determined which type of problem can be appropriate to solve with direct numerical simulation. After the direct numerical simulation studies, Reynolds averaged Navier Stokes models were researched. Reynolds averaged Navier Stokes models were examined under three subgroups which are single equation models, two equation models and other models. Prandtl's single equation model, Baldwin-Barth model and Spalart-Allmaras model were investigated under single equation models chapter. The models with two equations are divided into k-ε and k-ω models. Standard k-ε model, RNG k-ε model and realizable k-ε model are analyzed under this heading. Standard k-ω model, SST k-ω model and baseline k-ω model were examined under k-ω models chapter. The Reynolds strain model was the only model which belongs to other models heading related to the Reynolds averaged Navier-Stokes models. Also, It is spesified which types of flow problems are suitable for solving with these models. Following the studies about direct numerical simulation and Reynolds averaged Navier Stokes models, the large eddy simulation method was researched. This method can be represented as a combination of direct numerical simulation method and Reynolds averaged Navier Stokes methods. Smagorinsky-Lilly Eddy Viscosity Model, Dynamic Smagorinsky Model, WALE Model were examined seperately under the large eddy simulation heading. The specialities of these models, the way to determine the filter width, and the types of flows that are appropriate for them are explained in detail. In addition to this, it is also stated that what is the level of large eddy simulation method according to Reynolds averaged Navier Stokes methods and direct numerical simulation method in terms of computational cost and accuracy. Fluent and Cradle computer programs are used for computational fluid dynamics analysis. The version of Fluent program is 18.2. The Cradle program is also known as SC Flow which belongs to MSC company. As previously mentioned, turbulence models were analyzed in detail in order to make a proper selection in the analysis programs. However, although direct numerical simulation method and large eddy simulation method give more accurate results, it has been decided to carry out the analysis studies with Reynolds averaged Navier Stokes method considering computational cost. As a result of the studies about turbulence models, four different types of turbulence models were selected for using in computational fluid dynamics programs. The selected turbulence models are Baseline k-ω model, SST k-ω model, Realizable k-ε model and Reynolds stress model. It is thought that the open channel parts to be used in the flow experiments will be compatible with many turbulence models due to the low roughness ratio and the fact that the flow is not complicated. Therefore, when selecting turbulence models, attention was paid to their specific characteristics. In the selection of Baseline k-ω model and SST k-ω model, it is effective to use the k-ω equations in the regions close to the wall as the transport equations of the models and the k-ε equations out of this region. The purpoase of the selecting Realizable k-ε model is due to the fact that the model is the ultimate developed k-ε model and also provides more accurate solutions. The Reynolds stress model was chosen because of adding extra equations to the solution for modelling anisotropy. Experimental studies were conducted in parallel with the investigation of open channel flows and turbulence models. In experimental studies, the profiles of eight different flows were observed at four different parts at two different flow rates, and height measurements were taken at three different points from each flow. The cross-sectional types or curvature ratios of the open channel structures used in the experimental studies are different. The part used in the first experiment has a circular cross-section and the part is flat. The part used in the second experiment is close to the rectangular section and the part is flat. The part used in the third experiment has a circular cross section and the part used has a high curvature ratio. The part used in the fourth experiment has a circular cross section and the part used has a low curvature ratio. In the experimental setup, the slope of the replaced part is measured by the laser slope meter. The setting of two different flow rates is provided with frequency converter pump to the 25 and 40 Hz. After adjusting the frequency of the pump in order to avoid any error in the flow setting, the speed and flow rate of the fluid coming from the pipe have been measured by ultrasonic flowmeter and the frequency converter pump is working correctly. After these operations, height measurements were taken from predetermined points in each experiment. In the analysis studies carried out in Fluent, the flows in the experimental studies were modeled and similar measurements were taken at the measured points. These values were compared with the experimental results and the deviation amounts were calculated. Three-dimensional modeling was carried out in Fluent. The two-phase flow is defined using the VOF method. Transient flow properties were used. As a result, according to height measurements and analyzes, the closest result in Fluent program was obtained with SST k-ω model. It was thought that SST k- ω model would be suitable for analysis before comparison. The reason for this is that k-ω and k- ε are used together in a uniform way in this model. SST k-ω turbulence model was chosen as the main model. The purpose of the main model selection is to be used in the subsequent analysis studies. The results obtained from other models are very close to SST k-ω model. The reason for this is that when the flow profiles extracted by the models are examined, the flow profiles of the models are largely in agreement with those in the experimental studies, and the flow created in the experimental studies does not have a complex flow structure. In addition, the margin of error in the third experiment and the fourth experiments was higher. This may be due to the fact that the parts are not flat in the third and fourth experiments. Based on these experiments and analysis studies, it has been concluded that the flow in open channel flows can be modeled in computational fluid dynamics programs. In this way, analysis studies will be made for turbulence distributor systems to be developed. It can be considered as the systems that can be suggested as a result of the studies, to give air to the turbulence area or to modify this turbulent area by adding with a solid material.

Benzer Tezler

  1. Tez No

    397787

    Longitudinal dispersion of heavy particles in a free surface flow over rough bed

    Serbest yüzey akımında pürüzlü taban üzerinde ağır parçacıkların boyuna dispersiyonu

    SELÇUK DEMİRBAŞ

    Yüksek Lisans

    İngilizce

    İngilizce

    2015

    İnşaat Mühendisliğiİstanbul Teknik Üniversitesi

    Kıyı Bilimleri ve Mühendisliği Ana Bilim Dalı

    PROF. DR. MEHMET SEDAT KABDAŞLI

    DOÇ. DR. VEYSEL ŞADAN ÖZGÜR KIRCA

  2. Tez No

    421312

    Nükleer reaktör soğutucu kanallarında akışın Lattice-Boltzmann yöntemi ile benzeşimi

    Lattice-Boltzmann simulation of flow in nuclear reactor subchannels

    ALİ TİFTİKÇİ

    Doktora

    Türkçe

    Türkçe

    2016

    Mühendislik BilimleriHacettepe Üniversitesi

    Nükleer Enerji Mühendisliği Ana Bilim Dalı

    DOÇ. DR. CEMİL KOCAR

  3. Tez No

    28558

    Numerical computation of condensation rate of vapor on an open channel using k-E turbulence model

    Başlık çevirisi yok

    GÜRAY MOLLAOĞLU

    Yüksek Lisans

    İngilizce

    İngilizce

    1993

    Nükleer Mühendislikİstanbul Teknik Üniversitesi

    PROF. DR. HASBİ YAVUZ

  4. Tez No

    128953

    Pürüzsüz açık kanal akımlarında türbülanslı hız dağılımlarının sürtünme faktörüne etkisi

    The effects of turbulanced velocity distribution of smooth open channel flows on friction factor

    AHMET ALPER ÖNER

    Yüksek Lisans

    Türkçe

    Türkçe

    2002

    İnşaat MühendisliğiNiğde Üniversitesi

    İnşaat Mühendisliği Ana Bilim Dalı

    YRD. DOÇ. DR. AHMET BİLGİL

  5. Tez No

    33905

    Pürüzsüz açık konal akımlarında türbülanslı hız dağılımının laser doppler anemometresi ile incelenmesi

    Başlık çevirisi yok

    MEHMET ARDIÇLIOĞLU

    Doktora

    Türkçe

    Türkçe

    1994

    İnşaat MühendisliğiÇukurova Üniversitesi

    İnşaat Mühendisliği Ana Bilim Dalı

    PROF. DR. M. SALİH KIRKGÖZ

Geri Dön