Yağış simülatörlerinde verimliliğin arttrırılması: Performans parametreleri üzerine yeni bir yaklaşım
Enhancing efficiency in rainfall simulators: A novel approach to performance parameters
- Tez No: 994609
- Danışmanlar: DR. ÖĞR. ÜYESİ ERDAL KESGİN
- Tez Türü: Yüksek Lisans
- Konular: İnşaat Mühendisliği, Civil Engineering
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 2026
- Dil: Türkçe
- Üniversite: İstanbul Teknik Üniversitesi
- Enstitü: Lisansüstü Eğitim Enstitüsü
- Ana Bilim Dalı: İnşaat Mühendisliği Ana Bilim Dalı
- Bilim Dalı: Hidrolik ve Su Kaynakları Mühendisliği Bilim Dalı
- Sayfa Sayısı: Belirtilmemiş.
Özet
Yağış simülatörleri (YS), hidroloji, jeomorfoloji ve erozyon alanlarında çalışma yürüten araştırmacılar için vazgeçilmez bir araçtır; ancak, her araştırmacının kullandığı yöntemlerdeki farklılıklar nedeniyle performans parametreleri değerlendirmesi değişmektedir. Bu farklılığın temel nedenleri; örnekleme yoğunluğunun (kap sayısı), kapların yerleşim düzeninin, değerlendirme ölçeğinin (noktasal-alt alan-tüm alan) ve raporlama biçiminin (tek bir ortalama değer yerine dağılımın sunulması gibi) çalışmadan çalışmaya değişmesidir. Bu nedenle aynı simülatör için bile performans göstergeleri literatürde doğrudan karşılaştırılabilir biçimde raporlanamayabilmektedir. Bu çalışma, yağış parametrenin (yağış şiddeti, alansal üniformluk, yağmur damla çapı) ölçümünde mekansal değişkenlik kavramını incelemekte ve yağış simülatörleri için daha iyi bir kalibrasyon yöntemine katkıda bulunmayı amaçlamaktadır. Basınçlı bir YS, 2,7 m x 1,2 m (3,2 m²) test kanalında, 40-, 70- ve 100-mmsa-¹ sabit şiddetlerde kontrollü bir laboratuvar sisteminde simüle edildi. Kanal dokuz alt bölgeye ayrılmış ve 108-189-315 arasında değişen sayısı ve farklı konfigürasyonlarda (N1, N2 ve N3) deneyler gerçekleştirilmiştir. Böylece kap sayısı ve farklı dizilimlerinin yağış parametrelerinin belirlenme sürecindeki etkisi analiz edilmiştir. Noktasal yağış şiddetleri, kap dizilimine ve kap yerleşim konfigrasyonuna bağlı olarak 15 ila 185 mmsa-¹ arasında değiştiği gözlemlenmiştir. Bu durum, yağış paramtrelerinin tek bir ortalama değer ile ifade edilmesiyle, YS'de mevcut olan güçlü yerel değişkenliğin yakalanamayacağını belirlenmiştir. Christiansen Üniformluk (CU) katsayıları, alt alanlarda %66-96 arasında değişmesiyle belirgin bölgesel farklılıklar gösterdi; en iyi konfigürasyonda yerel CU katsayıları %96'ya ulaşırken, daha düşük yoğunluklu konfigürasyonlarda alt alan CU katsayı değerleri %66 kadar düşüktür. Kap sayısının artmasıyla ölçülen şiddetlerin yayılımı azlmış ve daha düzgün noktasal üniformluk desenleri oluşmuştur. Buna rağmen CU değerlerinde yalnızca küçük miktarda farklılıklar gözlemlenmiştir (tipik olarak belirli bir bit şiddet için ±%1 mertebesinde). Kap sayısıyla ilgili sistematik bir eğilim görülmemişitr. Yağmur damla çapı (RD) özellikleri, artan şiddetle birlike ortalama yağmur damlası çapında (D50) görece küçük ancak tutarlı bir artış gösterdi; 40 mmsa-¹'de 1,39 mm'den 100 mmsa-¹'de yaklaşık 1,60 mm'ye yükseldi ve daha yüksek şiddetli noktalarda daha fazla mekansal değişkenlik sergilemiştir. Özetlenen yöntem, yağış simülatörü performansını değerlendirmek için daha sağlam ve karşılaştırılabilir bir çerçeve sunarak, gelecek araştırmalarda (erozyon, sızma, yağış-akış vb.) simülatör tabanlı çalışmaların güvenilirliğini artıracağı öngörülmektedir.
Özet (Çeviri)
Rainfall simulators are common in laboratory hydrology and erosion studies since they can simulate rainfall events at predetermined intensities and durations, regardless of the supplies that nature would offer. This repeatability reduces experimental costs and time and allows the researcher to fix boundary conditions such as slope, soil texture, and antecedent moisture. However, the utility of simulator-derived results depends on the degree to which artificial rain resembles natural rain in those factors that drive runoff and soil detachment. In most studies, performance has been described principally by the plot-averaged rainfall intensity, and sometimes by a single plot-scale uniformity value. Practical as these measures are, they can overlook some critical realities. Pressurized nozzle systems often generate rainfall which is non-uniform over the plot due to nozzle type, the pattern of nozzle overlap, the height of the nozzles above the plot, pressure, and particularly the breakup of droplets in the air. Consequently, two experiments with the same plot average intensity may have quite different local rainfall. That can alter the infiltration-versus-runoff relationship, shift the infiltration thresholds at which runoff commences, and even create local erosion hot spots, thereby complicating comparative reasoning among studies. A second limitation is the prevalent omission of raindrop size distribution and kinetic energy. While intensity is often treated as the primary driver of runoff and erosion, splash detachment and particle entrainment depend more directly on droplet size, impact speed, and the kinetic energy flux delivered to the soil surface. These factors are not easy to measure, so many simulator studies appear very similar in terms of intensity and duration but vary greatly in the droplet regime and energy transfer. This inconsistency makes interstudy comparisons difficult and impedes progress toward standard reporting and calibration practices. It resolves these issues by developing an integrated, spatially explicit rainfall simulator performance evaluation approach. The methodology characterizes behavior at the plot level and within-plot variability for rainfall intensity and uniformity (Christiansen Uniformity Coefficient, CU), raindrop size distribution (with a focus on median volume diameter, D50), and kinetic energy for multiple target intensities. Additionally, there is an inquiry into how sampling density and layout impact the representation of spatial variability in an effort to make the rainfall simulations from the lab more interpretable and reproducible. Set up experiments, calibration, and sampling design: Laboratory rainfall simulator experiments were conducted on a rectangular channel 2.7 m long and 1.20 m wide. Spatially, the plot was divided into nine equal sub-areas (S1-S9) for spatial analysis. For this purpose, the platform holding the channel provided the possibility of setting the channel to a representative slope in order to better simulate real overland flow and couple it with hydrologic measurements. Rainfall was supplied by FullJet B1/GG-12W nozzles mounted en bloc at a fixed height above the channel. Nozzles were mounted to provide adequate overlap across the plot, since overlap geometry is the main factor influencing spatial patterns. System discharge and pressure had a wide tuning range. An empirical calibration linking discharge, Q, to plot-average rainfall intensity, I, was used to target three intensities: 40, 70, and 100 mmh-¹. During experiments, system stability was checked by flow and pressure gauage to keep desired conditions. Rainfall intensity was measured by collecting cups spread over the plot. The collected water was converted into intensity by using each cup's area and the sampling duration. Spatial uniformity was quantified with CU, which makes use of how each cup reading deviates from the plot mean. CU was calculated for both the full plot and for the sub-areas in order to spot local differences that are not visible in just one overall index. Accordingly, for the purpose of testing a sampling strategy, this was carried out under three cup-density scenarios going from coarse to dense-N1 to N3-along with three layout approaches: a regular grid layout, RGL; an centered-cluster layout, CCL; and an edge-cluster layout, ECL. These designs assess whether local extremes-like high-intensity overlaps or low-intensity edges-are captured under different practical measurement schemes. The size characteristics of rainfall were estimated using an indirect, laboratory-friendly methodology. For the drop size distribution, D50 was adopted as a measure to represent a robust indicator of the droplet regime. Results: Control of intensity, spatial variability and uniformity; the simulated plot-average intensities approached the target values of 40, 70, and 100 mmh-¹, thus confirming the adopted calibration procedure. However, the point-scale measurements revealed high spatial variability around all the targets. Usually, variability increased with increasing targets; this may indicate that more intensive operation emphasizes local heterogeneity in rainfall. From distribution-based analyses, it appeared that higher intensities have larger ranges and are more dispersed, implying that the difference between local and plot-average increases with intensity. Uniformity analysis added further insight: plot-average CU values were acceptable to good across intensities and tended to improve with higher intensity. However, sub-area CU showed that local non-uniformity could be persistent even when plot-scale CU looked decent. For instance, at 40 mmh⁻1, the plot-average CU was in the %70, whereas sub-areas varied between the %60 and above %90. At 70 mmh⁻¹, plot-average CU increased to the %70, with sub-areas varying more widely and reaching the %90 in the most uniform zones. At 100 mmh⁻1, plot-average CU was in the %80, while sub-areas formed a tighter, higher band. This suggests that a single plot-average CU can mask local pockets of low uniformity relevant to interpreting runoff initiation, soil sealing, and erosion. Drop regime and kinetic energy implications; D50 increased with intensity, from about 1.39 mm at the lowest target to about 1.60 mm at the highest. Since kinetic energy depends on both droplet mass and impact velocity (linked to terminal velocity), even modest shifts in D50 can noticeably change energy delivered to the surface. Thus, intensity alone doesn't fully describe erosive forcing: events with the same mean intensity can deliver different energy fluxes if their droplet regimes differ. Spatial variability in intensity and droplet properties also implies spatial variability in energy delivery, emphasizing that spatially resolved performance assessment is critical-especially for erosion and splash-related work. Impact of sampling density and layout; the sampling strategy strongly conditioned the characterization of rainfall patterns and performance metrics. Sparse sampling risks missing localized maxima and minima and inflates perceived uniformity, underrepresenting variability. Increasing cup density positively affected the resolving power of rainfall gradients and the identification of persistent zones but also increased workload. Comparisons between layouts revealed that the most informative layout depends on the objective of the study: center-cluster layouts capture the overlap zones in multi-nozzle setups, whereas edge-cluster layouts reveal boundary heterogeneity and low-intensity zones, and regular grids layouts balance coverage for contour mapping and sub-area summaries. These results suggest treating cup density and layout as core methodological choices and documenting them in simulator performance reports. Conclusions and recommendations: This work proposes a spatially explicit, multi-parameter framework for the evaluation of rainfall simulator performance. The key takeaways are: (1) hitting the target plot-average intensity is necessary but does not guarantee spatially consistent rainfall; (2) plot-average CU can mask local non-uniformity, and sub-area assessment improves interpretability; (3) droplet metrics and kinetic energy indicators are essential for a physically meaningful evaluation, as D50 and energy delivery change with operating targets; and (4) sampling density and layout influence inferred performance and should be chosen deliberately and reported transparently. In the light of these results, a comprehensive reporting framework for rainfall simulator experiments is suggested, which should comprise: i) target and actual mean intensity with statistics of variability, ii) CU at plot and sub-area scales, iii) explicit documentation of cup density and layout. This will allow for enhanced comparability, repeatability, and physical interpretability of laboratory rainfall simulations, particularly for erosion-focused and process-based hydrology studies.
Benzer Tezler
- Stewart platformu tasarımı
Design of the Stewart platform
BURAK ULAŞ
Yüksek Lisans
Türkçe
2009
Makine Mühendisliğiİstanbul Teknik ÜniversitesiMakine Mühendisliği Ana Bilim Dalı
YRD. DOÇ. DR. ZEKİ YAĞIZ BAYRAKTAROĞLU
- Arazi tipi değişken yağış yoğunluklu yağış simülatörü tasarlanması
Designing a variable rainfall density field type rainfall simulator
AYHAN HÜSNÜ GÜMÜŞ
Yüksek Lisans
Türkçe
2019
ZiraatTokat Gaziosmanpaşa ÜniversitesiBiyosistem Mühendisliği Ana Bilim Dalı
PROF. DR. ALİ KASAP
- A machine learning approach to forecast production performance of hydraulically fractured horizontal wells in tight oil sandstone reservoirs
Sıkı petrol kumtaşı rezervuarlarındaki hidrolik olarak çatlatılmış yatay kuyuların üretim performansını tahmin etmek için bir makine öğrenimi yaklaşımı
FERHAT YURTTAGÜL
Yüksek Lisans
İngilizce
2025
Petrol ve Doğal Gaz Mühendisliğiİstanbul Teknik ÜniversitesiPetrol ve Doğal Gaz Mühendisliği Ana Bilim Dalı
DR. ÖĞR. ÜYESİ İHSAN BURAK KULGA
- High precision motion control of mechatronic systems in presence of general uncertainties: Application with a heavy-duty parallel robot
Genel belirsizliklerin olduğu mekatronik sistemlerin yüksek hassasiyetle kontrolü: Paralel robot uygulaması
KAMİL VEDAT SANCAK
Doktora
İngilizce
2022
Makine Mühendisliğiİstanbul Teknik ÜniversitesiMakine Mühendisliği Ana Bilim Dalı
DOÇ. DR. ZEKİ YAĞIZ BAYRAKTAROĞLU
- Yağış verilerinin autorun analizi
Autorun analysis of precipitation data
KASIM KOÇAK
Yüksek Lisans
Türkçe
1992
Meteorolojiİstanbul Teknik ÜniversitesiMeteoroloji Mühendisliği Ana Bilim Dalı
PROF.DR. EREN OMAY